The 2026 World's Fair is completely sold out 🫡 ✅ The largest AI industry expo on earth ✅ Sold out on Leadership track for CTOs & VP AI's ✅ Sold out on Workshops tomorrow ✅ Sold out on ALL late bird tickets 🙌 65 side events still FREE all over SF (see website) What we will never sell out: Our commitment to publishing all the best AI engineering content for free online on YouTube. We have now opened limited overflow tickets for our expo and engineering tix — no seating guaranteed, sessions are first come first served. If you ARE one of our attendees, DO come down to Moscone for New Engineer Orientation tonight from 5p-9p to meet new friends and skip the morning crush for tomorrow. We expect EXTREMELY heavy last minute registration and need your help to load balance across days. Please give your speakers and sponsors all the love for all the effort they are putting into making this the greatest show we have ever done!

🆕 Releasing our entire RL + Reasoning track! featuring: • @willccbb, Prime Intellect • @GregKamradt, Arc Prize • @natolambert, AI2/Interconnects • @corbtt, OpenPipe • @achowdhery, Reflection • @ryanmart3n, Bespoke • @ChrSzegedy, Morph with special 3 hour workshop from: @danielhanchen of Unsloth! start here: piped.video/watch?v=PbHm2qKn… Happy weekend watching! and thanks to @OpenPipeAI for supporting and hosting this track!

The highly-anticipated MCP workshop with @MaheshMurag is now live! Join @AnthropicAI Applied AI Engineer (+ creator of MCP) @MaheshMurag in this full 2-hour workshop covering the Model Context Protocol, where he covers: 00:00 Introduction 00:35 What is MCP? 09:39 Building with MCP 26:25 MCP & Agents 1:13:15 What's next for MCP? YT version in next post.

🆕 Shipping our entire SWE Agents track! ft. - @cognition_labs Devin (@ScottWu46) - @GoogleDeepMind Jules (@rustinb) - @windsurf_ai Cascade (@kevinhou22 3x!) - @GitHub Copilot (@geektrainer) - @withGraphite Diamond (@TomasReimers) - @AnthropicAI Claude Code (@bcherny) - @Allhands_ai /OpenHands (@rbren_dev) - @imbue_ai Sculptor (@joshalbrecht) - @FactoryAI Droids (@EnoReyes) catch up on all the world's top coding agents including more workshops and talks from @augmentcode, @stanford, and more videos being uploaded!

🆕 Training Agentic Reasoners today's feature is @willccbb's triumphant return to the AIE stage RL track - now as part of @PrimeIntellect! A lot of agent builders are basically doing "RL by hand". He concisely explains current RL algorithms in one slide (!) but then argues that RL - particularly for open models - is stuck in math and code Q&A land the new hotness is multi-turn agentic RL, and the new verifiers library is the ultimate toolkit for building an agent and turning it into an RL loop. piped.video/watch?v=PbHm2qKn… More people should be exploring building better agent models and Will + PI is enabling that for everyone!

🆕 Releasing the entire AI PM track! Great products aren't vibe coded, they're carefully managed. We gathered the best PMs in the world to discuss how they build products that win. ft. - @raizamrtn (@gethuxe, fmr PM @notebooklm) - @bbalfour (Reforge, Hubspot) - @chrislovejoy_ (Head of Clinical AI, @anteriorai) - @benstein (Teammates, fmr CPO QuitCarbon/Arcadia) - @tommoor (@linear Head of Eng) - @auchenberg (Alloy, fmr PM Stripe, VS @Code) - James Lowe (Head of AI Engineering, UK Gov Digital Service) Send this to the PMs in your life!

Claude 3.5 Sonnet shows a 64% success rate on pull request evals. Claude 3 Opus was at 38% It's particularly strong in RAG and has near perfect recall over its 200k context window @alexalbert__ from @AnthropicAI

🆕 A 1 hour Masterclass on @aisdk, now free on YouTube for the first time ever! Covering: - Fundamentals of @Vercel's ultra popular AI SDK - Clone your own Deep Research in 30 mins! piped.video/watch?v=kDlqpN1J… Thanks to @nicoalbanese10 for the incredible work in rerecording his workshop given all the updates that have happened since NY!

Want to learn more about Context Engineering? Start here. 🆕 12-factor Agents: Patterns of reliable LLM applications piped.video/8kMaTybvDUw next featured talk on our Agent Reliability track is the fan favorite manifesto by @dexhorthy, who, among many other things, coined Context Engineering! Even if LLMs continue to get exponentially more powerful, there will be core engineering techniques that make LLM-powered software more reliable, more scalable, and easier to maintain. Factor 1: Natural Language to Tool Calls Factor 2: Own your prompts Factor 3: Own your context window Factor 4: Tools are just structured outputs Factor 5: Unify execution state and business state Factor 6: Launch/Pause/Resume with simple APIs Factor 7: Contact humans with tool calls Factor 8: Own your control flow Factor 9: Compact Errors into Context Window Factor 10: Small, Focused Agents Factor 11: Trigger from anywhere, meet users where they are Factor 12: Make your agent a stateless reducer

For the first time ever: @OpenAI's official definition of an agent for AI Engineers; plus case studies, design patterns, & lessons learned from building Swarm, Operator, Deep Research, and beyond. This is the multi-million dollar talk you don’t want to miss with advice typically reserved for their enterprise clients. Presented by @prashantmital, Member of Technical Staff at @OpenAI, & @tokisherbakov, Head of Solutions Architecture at @OpenAI. YT version in next post. Chapters: 00:00 Opener 00:17 Intro and overview 00:40 How OpenAI is structured 1:44 Enterprise AI customer journey 3:02 AI as part of your business strategy 6:54 Morgan Stanley case study 8:00 2025 is the year of agents 9:00 Defining Agents 9:54 Lessons OpenAI learned building Agents in the field 9:56 Lessons: Abstraction is a tool, not a crutch 11:43 Lessons:: Start Simple 13:00 Lessons: Use a network of agents for complex tasks 14:52 Lessons: Keep prompts simple; use guardrails to handle edge cases 16:05 Recap

Go beyond RAG without fine-tuning your model! In this talk from @klettphoebe of @NormalComputing, you'll learn about Extended Mind Transformers and how they select and attend to the most relevant information at each generation step. She demonstrates EMT's state-of-the-art performance and discusses important design decisions for long context applications. Preview below; full talk in next post!

Announcing Wave 1 of our keynote speakers! @gdb, President & co-founder @OpenAI @saranormous, founder @conviction @simonw, legendary AI Engineer @drperszyk, Cognitive Scientist @amazon AGI @hwchase17, CEO @langchain @solomonstre, CEO @dagger_io (& creator of @Docker) @jessemhan, CEO @morph_labs (ex GPT4) @_micah_h, CEO @ArtificialAnlys @sgrove, MTS, Alignment Reasoning @OpenAI Early Bird tix selling out very soon - AlE is 20 miniconferences all co-located in one place in SF so you can meet everyone who matters: - MCP - Tiny Teams (ARR > employees) - LLM RecSys - GraphRAG - Agent Reliability - Infrastructure - Product Management - Voice Al - Reasoning + RL - SWE-Agents - Evals - Retrieval + Search - Security - Generative Media - Design Engineering - Robotics/Autonomy And for the first time TWO two-day leadership tracks for CTOs and VPs of Al: Al in Fortune 500 and Al Architects Join us from Jun 3-5 in SF!

🆕 Releasing our entire Agent Reliability track! ft. - @tanmaigo (Hasura/PromptQL) - @dexhorthy (HumanLayer/Context Engineering) - @psomal (Temporal) - Anish Agarwal et al (@Traversal_ai) - @mr_cheu (Glean) - @calcsam (Mastra) - @itamar_mar (Qodo) Agents are very hyped, but not enough conferences have the focus and space to fully explore Reliable agents. Thanks to @PromptQL for helping create more reliable agents in the world. Enjoy!

We're excited to announce our speakers! CEO @Modular AI LEAD @MozillaAI ENG LEAD @OpenAI CEO @UnslothAI TBA @Microsoft TBA @AnthropicAI CEO @cognition_labs (Devin) CEO @anysphere (@cursor_ai) CTO @sourcegraph GEN AI @awscloud Product Lead @GoogleAI CEO @NousResearch VP AI Eng @Mastercard SVP AI @CapitalOne CEO @HypermodeAI CEO @echoai CEO @usebraintrust CEO @log10io CEO @neo4j CEO @DeepgramAI CEO @predictionguard CTO @FactoryAI CEO @convex CEO @HasuraHQ CPO @arizeai SR ENG @discord ENG @midjourney ENG @perplexity_ai VP PRODUCT @edgioinc CEO @crewAIInc CEO @OpenPipeAI CEO @BotDojo CEO @dottxtai CEO @e2b_dev DIRECTOR AI ENG @khanacademy CEO @lancedb ENG @character_ai CHIEF SCIENTIST @Gradient_AI_ VP PRODUCTS @Microsoft CTO @nixtlainc And many more, including more to be announced!

"Building Effective Agents" from @AnthropicAI has driven the agent engineering discussion ever since it launched. We're excited to air the first talk ever from @barry_zyj: piped.video/D7_ipDqhtwk ! The Core Ideas: - Don't Build Agents For Everything - Keep It Simple - Think Like Your Agents!

🆕 We're excited to release the first talk from the extremely popular MCP track — @chu_onthis, PM of MCP at @AnthropicAI, on: - the origin story of MCP - the evolution/updates of the MCP spec - (!!!) their REQUESTS FOR STARTUPS that people can build! Now live on the AIE youtube: piped.video/watch?v=x-8pBqWi… thanks for speaking, Theo!

Awesome demo showing the power of TypeScript and @typechat to get well-typed, structured data from LLMs 🤯 from @drosenwasser

LangGraph isn't just about creating agents; it's about crafting bespoke cognitive architectures tailored to specific business needs @hwchase17, CEO of @langchain

🆕 Structured communication is the bottleneck. Whoever communicates best becomes the most valuable programmer. Today's feature is The New Code, by @sgrove of @OpenAI! As LLMs increasingly do more and more code generation, the need for clear communication with Agents and with Humans goes *up*. The clearest artifact of communication is a specification, whether it is a PRD between PMs and Engineers, the US constitution between government and its people, OpenAI's model spec for LLMs. piped.video/watch?v=8rABwKRs… Sean applies the Model Spec to the GPT 4o glazegate issue - which actually preserved trust in the community that sycophancy was a bug! And since everything is a spec... we should think hard about how good specs are written, as we write them for our own AIs.

🆕 Releasing our entire Voice track! ft. - Sean Dubois, @DKundel (OpenAI) - Toki Sherbakov + Anoop Kotha (@OpenAI) - @kwindla + Chad Bailey (Daily) - @shresbm (Gemini @GoogleDeepMind) - Peter Bar (@Intercom) - @bnicholehopkins (Coval, ex Waymo) - @dwyer_neil (Gabber) - Jordan Dearsley (@Vapi_AI) - @tom_shapland (LiveKit) - @thorwebdev (ElevenLabs) - Rohit Talluri (AWS) + @arjundesai (Cartesia) Really honored to assemble the best Voices in Voice AI!

How do you build a @deepseek_ai R1 or @OpenAI Deep Research? @MorganStanley ML Researcher @willccbb tells us that it's largely due to Reinforcement Learning. In this talk, Will discusses how RL works and how to to think about it in the context of building agents. Preview here; full talks in following posts (X & YT).

Starting today, access to Replit AI is free. @amasad announced AI for all. "We'll be burning as much GPU as we're burning CPU."

We're excited to announce AI Engineer Paris! On Sept 23-24, our friends @gokoyeb are organizing the first-ever AIE conference in Europe. ai.engineer/paris We announced this event to our newsletter on July 4 and Super Early Bird sold out in 72 hours! Don't miss your chance to join us for this exciting inaugural event in Europe. Get tickets today, and we'll see you there!

We're excited to publish this talk from @thesephist, AI Lead @NotionHQ: The Hidden Life of Embeddings Get a preview of the talk in the interview below, & watch the full talk here 👉 ai.engineer/summit/schedule/…

Excited to announce @windsurf_ai as an expo partner! Windsurf is the first truly agentic AI-native IDE, capable of multi-step iterative reasoning and multi-file edits. Come learn from them at their booth -- & don't miss @kevinhou22's talk Feb 21!

The 1st project to come out of @mozilla builders is Llamafile The 2nd is sqlite-vec: Adding vector search to sqlite. Imagine a local LLM running on a Raspberry Pi with 0 internet connection

Imagine an agent in your IDE that is 10 - 20 steps ahead of you. It understands your entire codebase. It infers your needs & preferences. It scales with intelligence. In this talk from @kevinhou22, head of product eng @windsurf_ai, he explains that they are 90% there, and shows how Windsurf plans to get us to 100. Preview here; full talk in next 2 posts (X & YT).

Keynote & MCP Livestream: piped.video/watch?v=z4zXicOA…

We are now LIVE with the Agents track! piped.video/watch?v=D7BzTxVV… First talks from: - @swyx (Agent Engineering) - @sayashk (AI Agents That Matter) - @AarushSelvan (Deep Research) - @barry_zyj (Anthropic Agents) - @ZackRW (Sierra ADLC) - @willccbb (RL) Emcees: @Kseniase_ @nlw! Special opening number with @SunoMusic!

👀 Speaker drop! "The only conference your engineers NEEED to attend" the SOTA in AI Engineering from @saranormous @simonw @gdb @anthropicai @raizamrtn @ben_kus @ArtificialAnlys @jessemhan @thedanigrant @willccbb @JustinLin610 @shresbm @kwindla @ankrgyl (and a few more to be announced....)

The holy grail of Local AI: 🔓Open Source 🏃‍♀️Runs Fast 💻Runs on your machine 🌐Runs on all machines @Mozilla is leading the charge on the new local AI ecosystem: @jartine’s work on @llamafile and @stlhood announced sqlite-vec & $100k in non-dilutive funding for all local AI projects! Full talk in next post!

🆕 Releasing the entire MCP track! We curated the best talks and must-know topics in the MCP world from 1H 2025: - @chu_onthis, Anthropic: MCP's Origins + Startups - John Welsh, Anthropic: shipping Remote MCPs - @altryne (W&B) + @bhelx (Dylibso): Observable .tools for MCP - @digitarald (VSCode): Hidden Capabilities of the Full MCP Spec - @samuelcolvin (Pydantic): MCP is all you need - @jancurn (Apify): the rise of the Agentic Economy - @zeeg (Sentry): MCP is not good yet! plus: - Den + @Jujujuliakasper (Microsoft): MCP Auth - @labeveryday: Agents with Amazon Nova + MCP! - @danmason (Stride): Telemedicine Case Study and more to come! piped.video/playlist?list=PL… Enjoy!

Stop over-engineering agents Rahul Sengottuvelu started an AI Agent company in 2021, acquired by @tryramp where he's now head of Applied AI. In this talk, he presents alternative ideas for scaffolding agents, arguing systems that scale with compute beat systems that don't. YT version in next post

We're excited to announce the AI Engineer Code Summit! Read on for a bit of context; apply to attend at the link in our reply. AI coding tools and agents have redefined how we develop software. AI Engineering teams can move so fast that the message is clear: if you're not using AI code to build your software, then from the perspective of those who are using them, you're literally stuck in the past. This invite-only event brings together the top AI Engineers & AI Leaders who are actively using the latest in AI coding tools in their personal workflow and in their organizational structure. As users and innovators of frontier technologies & processes, we will connect to exchange stories, experience, and lessons with each other to drive our collective industry and profession forward. Anchoring the hallway track are talks from the top companies, AI Engineers, AI Architects, & founders/CEOs who will share their own experience, lessons, and advice for an advanced technical audience. Once again we have a "no vendor talks" policy to ensure maximum value transfer. Our stage will feature the top AI Coding builders from: - Google DeepMind - Anthropic - OpenAI - Amazon - Microsoft - Cursor - Cognition - Factory - Conductor - Augment Code - Amp Code - Cline - Goldman Sachs - McKinsey - MIT - Stanford - Cornell - & more! CFP also open until Sept 15, see link in reply. Additionally, the AI Engineer Expo will not only feature the top companies who are supplying AI Engineers with the latest tools, agents, and workflows for AI Code, but will also feature a diverse group of companies innovating in other areas. This expo will be a tour de force of productivity and innovation where the bleeding edge meets companies who have proved (or are forging!) product/market fit. Just like our February summit on "Agents at Work", this event has 2 tracks: AI Engineering (hands-on-keyboard) and AI Leadership (AI Architects, VPs, & other AI executives). These tracks are held on 2 separate days -- invitees can purchase one or both tracks. Summit is our highest-signal, invite-only event. Attendance is limited to 500 per track. Apply today at the link in the replies to connect with your industry peers and leaders to take your personal and company code to the next level.

We're still so early in AI. We're tacking it onto existing products The next Snapchats and Ubers will come from rethinking our apps from the ground up with AI @alexalbert__ , Head of DevRel @AnthropicAI

❤️

Big news! We're thrilled to announce the inaugural AI Engineer World's Fair, coming to San Francisco in March/April 2024 🎉 Cutting edge demos, products, & infra from 100+ companies, 50+ speakers across multiple tracks, & the most high signal conversations & connections of the year. Use code AIE for 25% off the (fully refundable) blind bird tickets, available now! Discount good for this week only. Get yours today and join us! 👇 ti.to/software-3/ai-engineer…

🆕 Full VOD for Jerry's talk is now available! Key insight: Every AI B2B SaaS basically boils down to building assistants and automation agents. to do this well, EVERY agent will need: - a document toolbox (preprocessing layer) (today, a document MCP server that can do file lookup, structured querying, retrieval, and manipulation, including over complex pdfs - using VLMs like @llama_index's LlamaParse — now with Excel support!!! the database of champions) - Agent orchestration (constrained vs unconstrained) or assistant vs automation/agentic workflows: and then putting the various patterns together for customer's specific situations --- FULL TALK HERE --- piped.video/watch?v=jVGCulhB… --- FULL TALK HERE --- @jerryjliu0 has been one of our top speakers for 2 years in a row - can he make it a hat trick without his iconic bucket hat??

🆕 2025 in LLMs so far, illustrated by Pelicans on Bicycles piped.video/watch?v=YpY83-kA… @simonw's keynote won our 2025 overall Best Speaker trophy, and was so well deserved! listen to the greatest highlights of 2025 in LLMs, Simon's explanation of the Lethal Trifecta in LLM security, and the release of PelicanBench!

Excited to announce our AIE CODE speakers! @maheshmurag @barry_zyj, Anthropic @bfioca @realchillben, OpenAI Codex @leerob, Cursor VP DX @eisokant, Poolside Co-CEO @simpsoka, Google Labs Director @jacob_d_kahn, Meta Codegen @StringChaos, Cursor & TerminalBench Al Harris, @amazon Kiro Lead @EnoReyes, Factory CTO Cofounders, @AppliedCompute Prime @willccbb @dexhorthy, Humanlayer, Context Eng coiner @aparnadhinak Arize @pashmerepat, Cline @joel_bkr, METR tba, @GoldmanSachs @nayshins, Netflix @nserrino, Gimlet Labs @ArmanHezarkhani, Tenex @thekitze, Sizzy Led by our wonderful emcee, @jedborovik of @julesagent. See you in NYC, Nov 20-22! (last round applications now open on our website)

🔴 Livestream is kicking off! piped.video/watch?v=5zE2sMka…

Announcing our speakers for the Reasoning + RL Track! ⚠️PSA: Tix nearly sold out, get em here: ti.to/software-3/ai-engineer… An incredible lineup of speakers at this track, featuring: @jack_w_rae, Principal Scientist @GoogleDeepMind @chrszegedy, former co-founder @xai @willccbb, Research engineer @PrimeIntellect @GregKamradt, President @arcprize @corbtt, co-founder & CEO of @OpenPipeAI @jessemhan, founder & CEO of @morph_labs @natolambert, Senior Research Scientist at @allen_ai & Founder of @interconnectsai @danielhanchen, CEO of @UnslothAI This track is also sponsored by @OpenPipeAI, who will be facilitating the track intros & speaker introductions as the in-room MC! World's Fair is 18 miniconferences all co-located in one place in SF so you can meet everyone who matters: - MCP - Tiny Teams (ARR > employees) - LLM RecSys - GraphRAG - Agent Reliability - Infrastructure - Product Management - Voice Al - Reasoning + RL - SWE-Agents - Evals - Retrieval + Search - Security - Generative Media - Design Engineering - Robotics/Autonomy And for the first time TWO two-day leadership tracks for CTOs and VPs of Al: Al in Fortune 500 and Al Architects Join us from Jun 3-5 in SF!

We're excited to publish @Wattenberger's talk, Climbing the Ladder of Abstraction She answers the question, "How might we use AI to build products focused not just on working faster, but on transforming how we work?" Preview below, full talk here: ai.engineer/summit/schedule/…

Claude 3.5 Sonnet was the darling of hallway track convos at World's Fair. So what's next for @AnthropicAI? @alexalbert__ tells us to build with the expectations that models will become smarter, cheaper, & faster in months, not years. piped.video/watch?v=EuC1GWhQ… See everything he announced in his full talk here!

Did you know that you can run C (and JavaScript, and even PHP) in Code Interpreter? @simonw demonstrates the incredible capability of Code Interpreter to run code that it was never supposed to run.

Deploying LLM apps to production comes with many challenges. @jerryjliu0 is the co-founder & CEO of @llama_index, and they've curated a list of best practices to help developers iterate on & develop production RAG & LLM apps. Watch the full talk here: ai.engineer/summit/schedule/…

10 Lessons on RAG Agents in Production from @douwekiela, creator of RAG @ContextualAI : piped.video/kPL-6-9MVyA 1. Better Systems > Better Models 2. Expertise/Specialization > AGI 3. Getting AI to work on noisy data > cleaning data 4. Pilot <----huuuge gap----> Production 5. Speed to market >>> Perfection 6. "Boring stuff" > Biz priorities 7. Easy to consume/integrate >>> 8. Wow your users! 9. Observability >> Accuracy (!) 10. Be Ambitious Bonus: Moravec's paradox in RAG: Complex synthesis is easy, but putting things in the right context is hard

🆕 For the first time: @gdb on the AIE stage! piped.video/watch?v=avWhreBU… Jensen: "How will AI Engineer workflows and pipelines change as more AGI comes online?" Greg: "I think the evidence has been shifting towards a menagerie of different models... there's actually a lot of power to be had by models that are able to use other models." We're so grateful for Greg to share his thoughts with the AI Engineer industry — but we're not done! pls comment/reply with followup questions for his @latentspacepod!

Announcing our speakers for the MCP Track! An MCP conference at the heart of the AI Engineer World's Fair, featuring: @JeromeSwannack, MCP core team @AnthropicAI @MaheshMurag, MCP core team @AnthropicAI @samuelcolvin, CEO @pydantic / @pydanti_ai @nilslice, CEO @dylibso + co-creator of Observable .tools @altryne, AI Evangelist @weights_biases + co-creator of Observable .tools @jancurn, founder & CEO of @apify @digitarald, Principal PM @code, @Microsoft Early Bird tix selling out very soon! Get your ticket here today (early bird closes Wed): ti.to/software-3/ai-engineer… AlE is 18 miniconferences all co-located in one place in SF so you can meet everyone who matters: - MCP - Tiny Teams (ARR > employees) - LLM RecSys - GraphRAG - Agent Reliability - Infrastructure - Product Management - Voice Al - Reasoning + RL - SWE-Agents - Evals - Retrieval + Search - Security - Generative Media - Design Engineering - Robotics/Autonomy And for the first time TWO two-day leadership tracks for CTOs and VPs of Al: Al in Fortune 500 and Al Architects Join us from Jun 3-5 in SF!

🆕Today's featured talk: @beyang on The emerging skillset of wielding coding agents! piped.video/F_RyElT_gJk It's raining coding agents! But while many are saying they're feeling the AGI, others say they're not that useful for serious programming. How much is hype and how much is a skill issue? We'll share empirical observations that help explain the divergence of developer opinion. And we'll cover emergent strategies uncovered by users of @AmpCode, a new coding agent in research preview, that can help you employ agents to complete more complex tasks in production codebases.

🆕 Releasing our entire Design Engineering track! ft: - @JohnPhamous, SFCompute - @vykthur, Microsoft/Autogen - @steveruizok, tldraw - @Vjeux, Excalidraw - @MaximillianNYC, Yutori - @SQuoraishee, NYT Games -@cwattrus, Flatfile LLMs today are 100x better than they were a year ago, but design thinking around the UX of AI has barely budged from ChatGPT and Canvas. We have gathered the top designers and design-engineers in AI to showcase their work. Enjoy!

From spreadsheets to Airbnb listings, the ability to 'zoom out' using AI lets us quickly understand and act. It's not just about more data; it's about clearer, concise, actionable insights. A look into what zooming out could look like from @Wattenberger and @AdeptAILabs

Hello, I’m an AI Agent and I run a brand account. AMA.

Announcing our Evals track speakers! Evals are the starting point of all production AI Engineering - when exciting demos turn into serious tools and agents. Our evals track is led by @ankrgyl, who had one of the top @latentspacepod episodes of the past talking about the evolution of evals at top AI Engineering companies like @Stripe, @Notion, and @v0. we're also excited to welcome: - @lateinteraction of @DSPyOSS - @rafalwilinski @vitorbal of @zapier - @ido_pesok of @v0 - @johnpdickerson CEO @MozillaAI - @jeffreyhuber CEO @trychroma - @jxnlco for the 3rd year in a row! thanks to @braintrustdata's Ornella for working with us to put this incredibly stacked lineup of the best Evals practitioners in the world!

Day 1 Livestream starts Monday, Oct 9 at 2:55pm PT Set a reminder & watch free here 👇 piped.video/watch?v=veShHxQY…

🆕 Second talk on the MCP track comes from John Welsh, MTS at Anthropic, on shipping Remote MCP support, covering what you need for authentication, security, and best practices to design your remote MCPs for the Pit of Success! piped.video/watch?v=0NHCyq8b…

a lot of CEO's ask "ok how do we know our AI investment is paying off?" - is AI only for greenfield projects? - where are the Fortune 500 AI stories? - Our developers are vibing... but are they just taking longer breaks or what? piped.video/watch?v=UXOLprPv… @beyang's story with @brunopassos is spectacular - using @sourcegraph to automate @bookingcom's massive, ancient codebase through 14 years worth of migrations for >3000 developers(!!)

Lessons from a year of building with LLM's In this special keynote from six industry heavyweights, they break down their popular white paper and offer key insights in 3 parts: piped.video/watch?v=qBHfQT3Y… 1⃣Strategic @BEBischof & @charles_irl discuss how you look to the future, see around corners, make big decisions 2⃣Operational @jxnlco & @HamelHusain discuss processes, teams, and workflows 3⃣Tactical @sh_reya & @eugeneyan discuss the specific techniques, tactics, and moves that have stood the test of 1 year’s time Watch their full talk here! 👇

Introducing Hyperspace: A new frontier in decentralized AI. Built on @HyperspaceAI's community-powered network, it combines prompt engineering, visual react flow, Python execution, and web browsing capabilities

📣Announcing the AI Engineer Summit Sponsor Expo! Join us to mingle with the founders and engineers at some of the most consequential companies in the world of AI. Special thanks to our Presenting Sponsor @Auto_GPT & our Diamond sponsors @FixieAI & @supabase!

welcome everyone! reply here with a short self intro (what you're looking for + what you can help with) if attending IN PERSON and open to DM's --- for IN PERSON attendees we STRONGLY recommend picking up your badges the night before or <1hr before the show starts. there will be a HUGE line at 7.30-9am on Wednesday - try to get ahead of it! enjoy and see you in the Slack, Conference App or IRL! tweet photos here!

Who's speaking at Summit on Feb 20 - 22 in NY? 👀 Member of Technical Staff, @OpenAI Staff ML SWE / TLM, @GoogleDeepMind Research scientist, @metaai Co-Creator, @PyTorch Member of Technical Staff, @AnthropicAI Product Manager, @GeminiApp Senior Software Engineer, @MethodFinancial Founder & CEO, @OpenPipeAI Head of Product Engineering, @windsurf_ai Vice President - Aladdin AI Engineering, @BlackRock Founder & CEO, @brightwaveio Author, AI Snake Oil Founder, @TakeoffAI Software Engineer, @JaneStreetGroup Staff Software Engineer, @the_superbill Head of Applied AI, @tryramp Co-Founder & CEO, @bee__computer CTO, @bee__computer Head of AI Engineering, @Bloomberg GTM - Enterprise, @AnthropicAI Partner, @Lux_Capital Founder, Parlance Labs Co-Founder, @specstoryai Cofounder and CTO, @privacera CoFounder, @augmentcode Managing Director, Product, @SignalFire Engineering Manager, Generative AI Foundations, @LinkedIn CEO & Co-Founder, @ContextualAI Director of Engineering - TR Labs, @thomsonreuters Founder & CPO, @arizeai Technical Lead, @AristaNetworks Director of Engineering/AI, @datadoghq VP of DevRel, @neo4j Senior Director, Operations & Insights, @pfizer Cofounder and CTO, @Sourcegraph Developer Experience & GenAI Innovation Lead, @bookingcom CoFounder & CTO, @Get_Writer Applied AI Engineer, @OpenAI Member of Technical Staff, @AnthropicAI Sr. Associate, Games, @solana AI Engineer, @Canada Co-Founder & CTO, @UseAvenue AI Engineer, @langchain Chief Llama Officer, @lifeatgoogle Founder, @pydantic DX, @amazon AI Evangelist, @weights_biases DX, @neondatabase Co-Founder & CEO, @Letta_AI Principal Developer Advocate, AI/M, @amazon AI SDK Maintainer, @vercel Principal, Head of AI Practice, @StrideBuild Research Engineer, 567 Labs And more!

Announcing our speakers for the SWE Agents track! Speakers include: @gdb, President & co-founder, @OpenAI @ScottWu46, CEO, @cognition_labs @bcherny, Claude Code, @AnthropicAI @rustinb, Product Manager, AI Coding, @GoogleLabs @TomasReimers, CPO & Co-founder, @withgraphite @ericdahoe, Member of Technical Staff, @augmentcode @itamar_mar, CEO, @QodoAI @k_dumez, DevRel Lead, @withgraphite Anish Agarwal, CEO and Co-founder, Traversal @EnoReyes, CTO, @FactoryAI Matthew Schoenbauer, Founding Engineer, Traversal Raj Agrawal, CTO, Cofounder, Traversal Raaz Dwivedi, Chief Scientist, Traversal @joshalbrecht, CTO and Co-founder, @imbue_ai @kylepenfound, Solutions Engineer, @dagger_io @jpadamspdx, Head of Ecosystem, @dagger_io @sam_alba, Co-Founder, @dagger_io @rbren_dev, CEO, @allhands_ai @MishaLaskin, CEO & co-founder, @reflection_ai World's Fair is 18 miniconferences all co-located in one place in SF so you can meet everyone who matters: - MCP - Tiny Teams (ARR > employees) - LLM RecSys - GraphRAG - Agent Reliability - Infrastructure - Product Management - Voice Al - Reasoning + RL - SWE-Agents - Evals - Retrieval + Search - Security - Generative Media - Design Engineering - Robotics/Autonomy And for the first time TWO two-day leadership tracks for CTOs and VPs of Al: Al in Fortune 500 and Al Architects Join us from Jun 3-5 in SF!

How to construct domain-specific LLM evaluation systems A talk by @HamelHusain, LLM consultant at Parlance Labs, and Emil Sedgh, CTO of @rechathq piped.video/watch?v=eLXF0Voj…

.@gdb gets surprise question from Jensen during his fireside with @swyx!

ChatGPT excels at general queries but struggles with specific, niche topics Even for subjects in its training data, it may fall short. It's important to set realistic expectations and pick the right tool for the right job Especially when coding - @ykdojo

A spicy take from @nutlope on shipping quickly and not getting bogged down in the details. Build, ship, iterate - and if it doesn't work out, pivot!

🆕 Website for the AI Engineer Summit NY: fresh off the press! ai.engineer/ Which speakers are you most excited to see?

- 120,000 players per week - 7 requests per second - Using OpenAI Clip Lessons learned in multimodality by @josephofiowa

Day 2 is going strong! Be sure to check out @jxnlco's talk at 2:50 PM**. Pydantic is STILL all you need

"If you look at 20 developers, 19 still don't use code AI" - @sqs, CEO of @sourcegraph

We're excited to announce @mckaywrigley as a speaker at our Agent Engineering track! Mckay builds in public with AI and is going to share some of his secrets with in-person attendees. From his abstract on our schedule (link in next post): I built a system where AI agents handle almost everything - from software development to customer support to marketing - to build, iterate, and maintain a live product. After 5 iterations, my most successful AI-built product was able to hit $3k MRR in under 2 weeks. Join him on Feb 21 in NY!

Can you use agents to build an agent company? @crewAIInc is a production-ready library to build & orchestrate multi-agent automations. @joaomdmoura, its founder & CEO, makes a number of announcements in this 13-minute talk! See link in next post for the full talk.

🆕 Building voice agents with OpenAI @DKundel launches the @openai Agents SDK in TypeScript -AND- gives the ultimate state of the art in best practices building voice agents with the latest S2S models! piped.video/watch?v=iXhba366…

Tinkering with side projects & Claude 3.5 Sonnet this weekend? In this 90-min workshop, @zswitten (staff prompt eng @AnthropicAI) uses prompts from attendees & takes them to the next level. It's truly a prompt masterclass. Full workshop now available👇 piped.video/watch?v=hkhDdcM5…

The AI Engineer World's Fair is confirmed for June 25 - 27 in San Francisco! Super Early Bird Tickets now on sale! Lock in the best rate by purchasing your ticket today, and use code TWEETY for an extra $100 off! Link in the next post🐦 Some more details about the event: Building on the success of Summit 2023 Last October, over 500 AI Engineers, founders, & partner companies came together for 3 days in San Francisco to learn about the latest advancements in AI Engineering. With over 5,000 applications and only 500 seats, this sold-out event exceeded our expectations in every possible way. We curated over 50 sessions across 3 days from engineers & founders representing companies including @AdeptAILabs, @amazon, @AnthropicAI, @Auto_GPT, @trychroma, @FixieAI, @github, @langchain, @llama_index, @CodiumAI, @microsoft / @msft4startups, @guardrails_ai, @DataStax, @newcomputer, @NotionHQ, @OpenAI, @PrefectIO, @Replit, @roboflow, @supabase, @getlindy, @_hex_tech, @vercel, @SmolModels and more. The result was so much learning, connecting, inspiration, and fun! All the talks are available on our YouTube channel for free: link in replies below. An Iconic Venue for World's Fair We have a full campus buyout of the beautiful Fort Mason Center in San Francisco. If you've never been, it features 2 large piers including one that just completed over $40mm in renovations. This newly-renovated pier, the Gateway Pavilion, is where the expo will be, across 2 floors, along with over 12 breakouts rooms for presentations, workshops, and discussions! This is especially exciting & valuable since our exhibitors have proven product-market fit with their offerings, and make your life as an engineer and business owner easier and more productive. So to be inclusive as possible, we have a limited number of "Expo Explorer" tickets at just $199 that gets you access to the Expo + the Expo sessions. Use code XTWITTER for an exclusive 50% off the Expo Explorer ticket through the end of February! Companies interested in sponsoring should email sponsors@ai.engineer for the sponsor guide. Speakers & Instructors who educate & inspire -- and who actually ship The pier next door (2 min walk) will host the General Sessions for 1,500+ people, breakout sessions and premium workshops, and a central cafe & lounge for full ticketholders with all day f&b, including breakfast & lunch buffets. Additional buildings will host workshops & breakout sessions (Gallery 308 for 500 people, Cowell Theater for 430 people) and the Presenting Sponsor's exclusive workshops & sessions (Store House). The three main stages and the 12+ breakout stages will feature over 100 sessions across seven tracks: 1. Multi-modal AI 2. RAG (Frameworks, Devtools, LLMOps) 3. Open Source Models 4. AI in the Fortune 500 5. CodeGen and Dev Tools 6. GPUs and Hardware 7. Autonomous Agents These tracks include talks, workshops, and facilitated discussion breakouts. All talks feature builders who are innovating in the exciting world of AI Engineering, across the stack and these 7 tracks. You can propose a session at our CFP, link in reply below. A special track for VPs of AI, Engineering, or Product New for Worlds' Fair is a VP ticket that gets you exclusive access to the General's Residence, a historic mansion perched on top of the park overlooking the Bay and the lower Fort Mason event center. Featuring 2 large rooms for breakout sessions, discussions, and networking events, plus many rooms with cozy lounge furniture and actual fireplaces and a beautiful lawn overlooking the Bay, you will connect & discuss with other company leaders in addition to the speakers, workshop instructors, and top sponsor company representatives. You also get access to the VIP welcome reception with speakers and top sponsors. We only have 200 VP tickets available, so don't delay. Our recommendation is to send 1 VP with at least 1 engineer attending the workshops and the most relevant sessions for your company. Group discounts available if you'd like to send more engineers. Four distinct ticket types We've reached our image limit for this post, but see the next reply for an image of what's included in each ticket type. Apart from the content, you'll be surrounded by other innovative technical builders. Every interaction will be valuable for you and your company. We look forward to seeing you at the biggest and most exciting AI Engineering event of the year! DMs open or email info@ai.engineer with any questions. If you'd like learn about sponsorships & see our incredible turnkey booths, email sponsors@ai.engineer for the sponsor guide. - Ben & Swyx

We are now live with the Leadership track! First talks from: @graceisford of @Lux_Capital @HamelHusain of Parlance Labs @gregce10 of @specstoryai @BoscoDurai of @privacera @colin__flaherty of @augmentcode Special opening number with @SunoMusic piped.video/watch?v=L89GzWEI…

We're excited to publish this talk from @kiwicopple, Supabase Vector: The Postgres Vector database Paul makes the case for pgvector and why every month, thousands of new AI applications are launched on @supabase, powered by pgvector. Full talk here 👇 ai.engineer/summit/schedule/…

breaking news: $NVDA loses $56 BILLION in market cap after @jartine @stlhood talk at AIEWF

~2 hour MCP workshop from @MaheshMurag, Member of Technical Staff @AnthropicAI piped.video/watch?v=kQmXtrmQ… This workshop talks about the philosophy behind Anthropic's Model Context Protocol, its impact on the broader ecosystem since launch, and how developers can use it to build context-rich AI apps and agentic experiences.

What is the future we're building towards? @romainhuet gives us a glimpse into what @OpenAI built into GPT-4o. His live demo shows how native multimodality capabilities reason across text, audio, and vision in real time. The speed at which this takes place (including interruptions) is particularly fascinating! See his chat with @SarahChieng for a glimpse into his talk. Link to full talk in next post!

Our full schedule is now live! Which are you most looking forward to? ai.engineer/summit/schedule

Photos from the event are starting to drop! These are the "teasers". Our photographer is hard at work sorting through the 38,000 photos taken to get us a manageable folder 😂 We'll post the album once it's ready!

Congrats to @openrouter for the $40m fundraise! We're excited to publish OpenRouter's first ever keynote about the origin story and coming roadmap: piped.video/84Vtz2IL1Ug Timestamps: Initial Question [01:16]: The story begins in early 2023 with the founder, Alex Atallah, pondering if the AI inference market would be dominated by a single player. He noticed the emergence of new models beyond OpenAI and a growing desire from developers to understand the nuances of different models, including their moderation policies [01:48]. The Rise of Open Source [02:35]: The video highlights the beginning of the open-source AI race, with early models like Bloom 176B and OPT from Facebook [02:46]. A pivotal moment was the release of Meta's Llama 1 in February, which surprisingly outperformed GPT-3 on many benchmarks [03:28], signaling a shift in the landscape. The Alpaca Moment [04:38]: A major breakthrough occurred in March 2023 with the distillation of Alpaca. Stanford researchers demonstrated that by fine-tuning Llama 1 with outputs from GPT-3, they could transfer the style and knowledge of a larger model to a smaller one for less than $600. This proved that creating powerful, specialized models no longer required massive budgets [04:58]. From a Chrome Extension to a Marketplace Window AI [06:43]: Before OpenRouter, Atallah launched Window AI, an open-source Chrome extension that empowered users to select their preferred LLM for any web application. This project laid the groundwork for what was to come. The Launch of OpenRouter [07:18]: OpenRouter was co-founded with Lewis, the creator of the framework that Window AI was built on. Initially, it was a simple aggregator to collect models in one place. Growth and Evolution [07:57]: OpenRouter quickly evolved into a marketplace, driven by the proliferation of model providers with varying prices, performance, and features. The platform has seen impressive growth, with a 10-100% month-over-month increase for two years. It now offers a single API for over 400 models from more than 60 providers [08:07]. Marketplace Dynamics [08:57]: The transition to a marketplace was a response to the complexity of the growing AI ecosystem. By aggregating providers, OpenRouter helps developers achieve better uptime for both open-source and closed-source models and provides valuable data on latency and throughput [10:27]. The Future of OpenRouter Expanding Modalities [17:02]: The future vision for OpenRouter includes incorporating models that can generate images and "transfusion models" that allow for conversations with images. Smarter Routing [17:51]: The platform plans to implement more sophisticated routing mechanisms, including geographical routing and enterprise-level optimizations for GPU allocation. Enhanced Discovery [18:07]: To help developers find the best models for their needs, OpenRouter aims to improve prompt observability, introduce more granular model categorization, and continue to offer competitive pricing.

This talk has already been seen by 13k people in just 3 days 🤯 Here's the interview with @soumithchintala where he describes the inspiration for his talk on building private, local AI agents.

🆕 We're excited to release the entire Tiny Teams track! ft. @SidBendre - Oleve cofounder @thisisgrantlee - Gamma CEO @EricSimons - Bolt CEO @nutlope - Together DDR @MaxBrodeurUrbas - Gumloop CEO @VikParuchuri - Datalab CEO @alxai_ - Every Head of AI > Tiny Teams are the rising trend of small but extremely productive teams augmented by agents...“teams with more m in ARR than employees”. We're writing the playbook as we speak!

We're excited to announce our Presenting Sponsor: @Microsoft! Engineers & @msft4startups team-members will be on-site for keynotes, workshops, discussions, demos, and perhaps some exciting announcements! Get your 🎟️ today & join them June 25-27 in SF! ai.engineer

.@danielhanchen used to work at NVIDIA making GPU algos go fast -- and also helped NASA engineers process data from a Mars rover faster! Today he's the co-founder of @UnslothAI, and their open source package makes finetuning of LLMs 2x faster and uses 70% less VRAM with no accuracy degradation. Come learn from him and meet him in SF on June 25 - 27!

We're pleased to start shipping talk video recordings! On each talk's page from the Summit website, you'll find: 📼 Talk Recording 🔡 Talk Transcript 📸 Speaker photos We aim to publish 2 - 3 per day, in order of appearance. First up: @Benghamine's opening welcome message, which also announces our event mobile app, Network, as a newly-open source project! Watch here 👇 ai.engineer/summit/schedule/…

Open Questions for AI Engineering @simonw asks the hard questions -- do you have the answers? This closing keynote recapped not only the event, but the questions that our industry needs answered to get us to the next step. Watch his full talk here 👇 ai.engineer/summit/schedule/…

Now live: @amasad & @pirroh in their keynote, Building AI for All, where they announce: ➡️@Replit is providing AI for their millions of users, for free! ➡️A new AI model! ➡️They open sourced their model! Watch/read the full talk & announcements here 👇 ai.engineer/summit/schedule/…

We're excited to publish @eugeneyan's talk! Drawing from academic research, industry resources, and practitioner know-how, he talks about practical patterns for integrating LLMs into systems and products. Watch here 👇 ai.engineer/summit/schedule/…

We're excited to announce our expo partners for Summit! Come meet the founders and leading engineers at these companies leading support & innovation in the world of AI Engineering: @solana @Sourcegraph @rungalileo @baseten @HasuraHQ @datadoghq @windsurf_ai @Get_Writer @weights_biases @ellipsis_dev @elevenlabs @gitpod @vellum_ai @LangbaseInc @PortkeyAI @daytonaio @trydaily

The calm before the storm #AIEsummit

Workshop Day going strong! Standing room only for @jneuwirth4 and @zswitten of @AnthropicAI

Next track we are releasing: the LLM RecSys track! Recommendation systems are one of the single most valuable applications of AI in the world, from LinkedIn to Netflix to YouTube to Instacart to Pinterest to Gemini to Amazon and even @elonmusk's @x feed via @TheGregYang. @eugeneyan curated some of the best LLM RecSys work in the world - many of these talks only accessible if you are a recsys industry insider - now for the first time public to all AI Engineers! Enjoy! piped.video/@aiDotEngineer/v…

how did pgvector and @supabase happen? it all started with an email! @ggrdson

"Agents... need SO MUCH of your Life Context that you're better off keeping them local and private." Our closing keynote speaker from @aiDotEngineer NYC: an absolute banger from @soumithchintala on why you need open source models like Llama 4 - though he acks that open models are way behind especially for **CATASTROPHIC ACTION CLASSIFIERS**, a key enabler for agents Personal Agents need all the context, but what models/companies would you trust with handing all information about you? How do you build an agent you trust as much as Gmail? or capture context as novel as the Meta Raybans? piped.video/watch?v=jMoAaZP_…

"The products you'll be able to build 6 months from now, a year from now... are going to be incredible" - Simón and Logan from @OpenAI

Official photos from the AI Engineer World's Fair 2025 are now available! Day 1: photos.google.com/share/AF1Q… Day 2: photos.google.com/share/AF1Q… Day 3: photos.google.com/share/AF1Q… Attendees can use our photos without attribution, although we'd appreciate it -- website url or tag us on social. All others: attribution required: on social tag and credit AI Engineer World's Fair, on blogs credit us and link to our website.

What RL Means for Agents - a talk by @willccbb, ML Researcher at @MorganStanley