NOV 10-12, 2026 // SF, CA

AI ENGINEER CODE SUMMIT 2026

Coding agents, code review, and software factories took the stage in San Francisco. Watch both days of talks or pick a session for its recording, illustrated essay, and transcript.


PRESENTED BYGoogle DeepMind//SUPPORTED BYG2i

AI Engineer CODE 2026 · Watch & read

The talks, in writing.

Rewatch both conference days, or browse individual sessions for their illustrated reading editions, recordings, and full timestamped transcripts. Separate recordings restore interrupted talks where available.

Explore all 50 sessions →

Browse the sessions

50 of 50 sessions · Event order · timestamps are positions in each source recording

Wednesday, November 11

Full livestream ↗

Cloudflare

From cloudflare is a neolab

November 11 · Main stage

cloudflare is a neolab

Michelle ChenCloudflareGroup Product Manager

Cloudflare built a distributed inference network around a bet: models would get smaller and faster, making the distance between the model and its user increasingly important.

SpaceXSI

From All Code Is Slop (Except Mine)

November 11 · Main stage

All Code Is Slop (Except Mine)

Lauren TanSpaceXSIPrincipal Engineer

When an agent makes the same mistake again, it is tempting to add another warning to AGENTS.md. Lauren Tan wants the correction to land somewhere harder to ignore: the codebase, an API, or an executable check.

Intrinsec AI

From Read the inference state before the agent acts

November 11 · Online track

Read the inference state before the agent acts

Speaker identity unverifiedIntrinsec AI

A guardrail can inspect a prompt before the model receives it and inspect a response after the model produces it. But between those checks, the model has already transformed its input into internal state. The talk asks whether an inference operator can use that state as an additional security signal.

Amazon Web Services

From Specs have bugs too

November 11 · Online track

Specs have bugs too

Rajath KotyalAmazon Web Services

An e-commerce specification says that an order should become a backorder when inventory is unavailable. Elsewhere, it says that no order should ever become a backorder.

Google DeepMind

From I deleted 90% of my harness (And It Got Better)

November 11 · Main stage · Excerpt

I deleted 90% of my harness (And It Got Better)

Philipp SchmidGoogle DeepMindMember of the Technical Staff

An agent can reach the right answer while taking an unnecessarily expensive route. Philipp Schmid's harness experiments focus on that difference: two systems may look equivalent in an accuracy table, yet consume very different numbers of turns, tokens, and dollars to finish the same work.

Sonar

From Building a Safe Software Factory

November 11 · Main stage

Building a Safe Software Factory

Andrew FinchSonarStaff Software Engineer

A coding agent makes it easier to produce a pull request. That does not settle whether the change should merge, whether it behaves correctly after deployment, or whether a vulnerability discovered next week changes the answer.

Cline

From The Improvement Loop Is the Product

November 11 · Online track

The Improvement Loop Is the Product

Ara KhanCline

A coding agent can help build the system that makes coding agents better. That observation is easy to turn into a story about a distant intelligence explosion. Ara Khan instead begins with a nearer question: what should an engineer work on when a new model can erase the value of a side project overnight?

Lovable

From What it takes to set intelligence free

November 11 · Main stage

What it takes to set intelligence free

Fabian HedinLovableCTO

Lovable nearly failed by building for agents too early. Fabian Hedin opens with that admission, not with the architecture diagram. In 2023, the team could see a future of agents delegating to agents, but the models available at the time could not reliably support the elaborate system they built around them.

Independent

From 2026 in LLMs

November 11 · Main stage

2026 in LLMs

Simon WillisonIndependentCreator of Datasette

Simon Willison organizes his year-in-review around three inflection points: coding agents becoming everyday tools, models becoming capable of pursuing much larger goals, and generated games becoming genuinely enjoyable to him.

Thursday, November 12

Full livestream ↗

OpenAI

From Ambition in AI

November 12 · Main stage

Ambition in AI

Jason LiuOpenAIDeveloper Experience

A thousand agents running in the background is an impressive quantity. It does not tell Jason Liu whether their owner is doing ambitious work. He wants to know what the person is trying to make, whether they care about the result, and whether they can recognize when it needs another pass.

Ayush Bhardwaj

From Your agent hates your CLI

November 12 · Online track

Your agent hates your CLI

Ayush Bhardwaj

One agent downloads a documentation page and keeps forty lines. Another parses a dependency report and keeps the tail because that is where the totals happen to live. A third runs a test suite and truncates the output. Each gets an answer by writing a small program around the program it was asked to use.

Cognition

From A New Age of Software Engineering

November 12 · Main stage

A New Age of Software Engineering

Walden YanCognitionChief Product Officer

Cognition's merged pull requests grew much faster than its headcount. Yan opens with that internal comparison because it creates a more interesting question than whether an agent can write code: what does an engineering organization have to change to absorb all that work?

Exa

From What Your AI Doesn't Know

November 12 · Main stage

What Your AI Doesn't Know

Jeffrey PinnerExaCTO

A model can reason well and still lack the fact that makes its answer correct. Pinner opens with the growth of search traffic at Exa, then turns to the more durable reason agents need search: the world changes after training, and training never contained all of the world in the first place.

AWS

From Ship More as a Team of One with Kiro Crew

November 12 · Main stage

Ship More as a Team of One with Kiro Crew

Doug ClausonAWSPrincipal Product Manager, Kiro

Few people raise their hands when Clauson asks whether their teams are getting everything they expected from AI. His diagnosis is that many teams have attached an agent to an existing process and accelerated only the code-generation step. The rest of software development still runs at its previous speed.

Google DeepMind

From Judgement is the last job

November 12 · Main stage

Judgement is the last job

Ishaan DeyGoogle DeepMindSoftware Engineer

When building a prototype becomes cheap, it changes when a team can learn. Instead of spending most of its effort imagining the right interface before implementation, the team can make several versions, try them, and discover details it could not have reasoned out in advance.

Factory

From A software factory needs pressure in both directions

November 12 · Main stage

A software factory needs pressure in both directions

Dexter HorthyFactoryMember of Technical Staff

A software factory already existed before coding agents. Someone identified a problem, translated it into work, built a change, ran tests, reviewed it, released it, and watched what happened. Dexter Horthy's argument is that replacing the builder does not make the rest of that system optional.

Independent

From Keep the human inside the creative loop

November 12 · Main stage

Keep the human inside the creative loop

Thariq ShihiparIndependent

Thariq Shihipar starts with predictions he got wrong. Agents did not settle into one permanently correct tool interface. Work he expected to stay inside familiar web frameworks expanded into other languages and applications. Capabilities kept moving beyond the boundary he had drawn around them.

//ABOUT THE SUMMIT

BUILT FOR PRACTITIONERS

Speakers shared the workflows they use to build, evaluate, and deploy coding agents.

THE RIGHT PEOPLE IN ONE ROOM

Top AI engineers building coding agents, founders and engineering leaders rolling out coding and internal agents, and researchers advancing AI software engineering.

THREE DAYS IN SAN FRANCISCO

A day of AIE workshops, two days of sessions, the Summit Expo, and plenty of breaks for the hallway track.

HIGHLY SELECTIVE

Attendance at the 2026 summit was by application.

//CODE 2027

AI Engineer CODE returns to San Francisco in 2027. Join the mailing list for dates and ticket announcements.

JOIN THE CODE 2027 MAILING LIST

//SPEAKERS

> Three themes anchor AIE CODE 2026: fighting slop in software factories, building with System One models, and rewriting critical apps.

Keynotes // by theme

Featured keynotes // 15 more on the main stage

Emcees // your hosts on stage

Talks & workshops // 69 in person

Show 51 more ↓

Online talks // 23 streamed

//PROGRAM

The 2026 program covered coding agents, code review, evaluations, and the infrastructure behind production workflows.

Explore more talks about coding agents →

Agenda overview // all times PT

Tue, Nov 10

Workshops & Welcome Reception

  • 8:00 AMBadge Pickup and Breakfast
  • 9:00 AM – 12:40 PMMorning Workshops Block
  • 9:00 AM – 12:40 PMWorkshop Talks Block
  • 11:10 – 11:40 AMMorning Break
  • 12:40 – 1:50 PMLunch
  • 1:50 – 5:30 PMAfternoon Workshops Block
  • 1:50 – 5:30 PMWorkshop Talks Block
  • 4:00 – 4:30 PMAfternoon Break
  • 5:30 – 8:30 PMWelcome Reception (Expo Hall)

Wed, Nov 11

Keynotes & Expo Talks

  • 8:00 AMBadge Pickup and Breakfast
  • 9:00 – 10:40 AMMorning Featured Keynotes
  • 10:40 – 11:20 AMMorning Break
  • 10:55 AM – 12:55 PMMorning Expo Talks
  • 11:20 AM – 12:55 PMMorning Keynotes
  • 12:55 – 2:05 PMLunch
  • 2:05 – 4:05 PMAfternoon Expo Talks
  • 2:05 – 3:40 PMAfternoon Keynotes
  • 3:40 – 4:20 PMAfternoon Break
  • 4:20 – 5:30 PMAfternoon Featured Keynotes

Thu, Nov 12

Keynotes, Expo Talks & Afterparty

  • 8:00 AMBadge Pickup and Breakfast
  • 9:00 – 10:40 AMMorning Featured Keynotes
  • 10:40 – 11:20 AMMorning Break
  • 10:55 AM – 12:55 PMMorning Expo Talks
  • 11:20 AM – 12:55 PMMorning Keynotes
  • 12:55 – 2:05 PMLunch
  • 2:05 – 4:05 PMAfternoon Expo Talks
  • 2:05 – 3:40 PMAfternoon Keynotes
  • 3:40 – 4:20 PMAfternoon Break
  • 4:20 – 5:30 PMAfternoon Featured Keynotes
  • 6:30 – 9:30 PMAfterparty

//FOR AGENTS & HACKERS

> The schedule and speakers are open data. Build your own schedule app, agent or tool and tag @aiDotEngineer.Same live data as the full schedule.

// QUICK START

# Day 1 (Nov 10) talks, straight from the live schedule
curl -s https://ai.engineer/code/2026/sessions.json \
  | jq '.sessions[] | select(.date == "2026-11-10") | {time, title, room}'

# Give your agent the whole event as context
curl -s https://ai.engineer/code/2026/llms-full.txt

# Add the AIE MCP server (Claude Code shown; any MCP client works)
claude mcp add --transport http aie https://ai.engineer/mcp

//SPONSORS

Thank you to the teams that supported AI Engineer CODE 2026.

// PRESENTING SPONSOR

Google DeepMind

The engine room of Google. Building AI safely and responsibly to solve the world’s most complex problems.

// OFFICIAL LOUNGE SPONSORS

Exclusive spaces to connect, recharge, and meet fellow builders, made possible by our lounge sponsors.

// PLATINUM

AWSCognitionDatabricksExaFactorySonar

// GOLD

AtlassianAuth0CiscoCodeRabbitMixedbreadNiteshiftOvercutParallelPineconeRenderSentryUnblocked

// SILVER

BANDCloudflareDatadogJudgment LabsKeycardNimbleParasailWarp

// SUPPORTING PARTNERS

FirecrawlVercelAgentsInTheCloudAmpAviatorBasetenComposiodbt LabsDiffusionE2BFivetranGimlet LabsGleanGmailGoodfireHigher Order CoHumanLayerLovableMonacoOsmanticRampReflectionUiPathXZoomInfo

Interested in sponsoring? Contact sponsorships@ai.engineer.

2026 SPONSOR GUIDE →

//FAQ