World Cup and Multi-Track Programming
What It Was
July 1 was the densest AI Engineer World's Fair 2026 programming day at Moscone West: a morning main-stage arc, a many-track conference grid, leadership-room sessions, expo-stage talks, and the $100,000 AIE Startup Battlefield. The day opened with Battlefield setup and then used the Main Stage to frame the technical thesis for the rest of the program. Barr Yaron set the day around the 2026 state of AI engineering, shifting attention from model fascination to the repeatable engineering practices implied by the previous year's results. John Ousterhout pushed the thesis down into transport infrastructure, arguing that modern inference workloads stress TCP and RDMA throughput assumptions and need designs such as Homa. Maxime Rivest and Isaac Miller made task definitions separable from model choice. Mike Krieger and swyx treated Anthropic's lab-building practice as an operating lesson, and Emil Eifrem closed the morning arc by casting agents as thinner clients on top of an ontology-based semantic layer.
From 10:45am onward the schedule split into a stack map rather than a single theme. Harness Engineering concentrated on production discipline for agents: tokens as work units, verifiers for computer-use agents, adversarial morality stress tests, reliable agent stacks, domain-agent cages, sandboxed bash execution, and skepticism about heavyweight agent frameworks. Agentic Engineering covered the tool layer around MCPs, CLIs, skills, auth, coding-agent exams and benchmarks, multiplayer development, and always-on production agents. The Inference track moved from distributed inference operations and policy-based routing to benchmark reliability, token demand, model ownership, scaling platforms, small-model clusters, and frontier inference clouds. The recurring engineering question was not whether agents could act, but how teams define the work, observe the run, bound the blast radius, and decide when the result is correct.
Several tracks treated agents as actors inside markets, enterprises, and regulated workflows rather than isolated demos. Agentic Commerce returned repeatedly to wallets, USDC and nanopayments, x402 apps, seller monetization, spending controls, payment infrastructure, and the practical problem of teaching agents to pay. The linked commerce talks make that concrete: multimodal shopping agents, machine-to-machine payment rails, agent wallets, and x402 app infrastructure all assumed agents would become buyers with permissions, budgets, and settlement paths. AI in Finance paired autonomous multi-agent research, money-aware models, event-sourced systems, verifiable financial AI, investor memo discipline, skills vetting, and the human bottlenecks that still shape finance-agent workflows. Healthcare sessions emphasized guardrails, ambient documentation, member-facing health AI, clinical intelligence, patient-scale deployment without A/B tests, gaps in generic voice stacks, X12 as an agent harness, and parallels between trading desks and clinical trials.
The leadership rooms extended those concerns into organization design. The CTO Circle session on agentic product development sat beside talks on global shipping agents, AI-native scaling across the org, enterprise contract evaluation, token FinOps, AI factory platforms, chip-design teams moving as one body, and agent enablement as a team capability. The Startup Battlefield added a parallel proof point: twenty AI startup founders pitched on the same day that the main program was asking how agentic products become durable businesses, not just impressive demos. The program therefore paired operating models with market pressure: product teams, platform teams, investors, and enterprise buyers were all treated as parts of the agent system.
The day also exposed the conference's synthesis layer. Graphs and ontology sessions treated memory, context, execution graphs, relational context engines, and systems of context as substrates for better agents, from CrabRAG's graph-memory framing to BabyAGI's active graph runtime and relational context engines pitched as a way to reduce token burn. GTM Engineering made buyer understanding, orchestration, recommendation loops, trust mechanisms, and knowledge systems feel like technical infrastructure. Local AI was framed through trust, control, optimization, edge compression, and frontier intelligence under a desk. Generative Media crossed model training, HTML as an agent-native interface, video editing, game engines, music, realtime video agents, and video memory for world-model training.
The expo stages sharpened the edge cases that the formal tracks kept circling: agents that cannot judge correctness, software factories that are not factories, hybrid retrieval beyond vectors, auth for agents, relational context to reduce token burn, local laptop security risks, OpenTelemetry tracing, vLLM speculative decoding on Blackwell, voice as an interface, microVM sandboxes, privacy gaps, verification gaps, self-modifying tools, and document-search systems judged by results rather than retrieval alone. Across the connected sessions, July 1's recurring claim was that AI engineering in 2026 is less about a single model and more about the surrounding system: task contracts, graph memory, policy routing, telemetry, sandboxes, payments, guardrails, cost attribution, and evaluation loops that make agentic work inspectable and controllable.
Scheduled Sessions
- 2026 07 01 session battlefield setup — 8:30am-10:30am · Startup Battlefield
- 2026 07 01 barr yaron the 2026 state of ai engineering — 9:00am-9:20am · Harness Engineering
- 2026 07 01 john ousterhout tcp and rdma are killing inference throughput homa can fix it — 9:20am-9:40am · Software Factories
- 2026 07 01 maxime rivest the unreasonable effectiveness of separating the task from the model — 9:40am-10:00am · Harness Engineering
- 2026 07 01 mike krieger how anthropic builds lessons from labs — 10:00am-10:20am · Harness Engineering
- 2026 07 01 emil eifrem thinner agents on a smarter substrate the ontology based semantic layer — 10:20am-10:30am · Graphs
- 2026 07 01 katelyn lesse tokens should have jobs — 10:45am-11:05am · Harness Engineering
- 2026 07 01 sangwu lee training krea 2 what matters in generative model training — 10:45am-11:05am · Generative Media
- 2026 07 01 nidhi kaushik vyas designing multimodal collaborative agents for next gen commerce — 10:45am-11:05am · Agentic Commerce
- 2026 07 01 brendan rappazzo alphalab autonomous multi agent research across optimization domains with frontier llms — 10:45am-11:05am · AI in Finance
- 2026 07 01 nader khalil state of the union why local why now — 10:45am-11:05am · Local AI
- 2026 07 01 stephen chin crabrag why automated assistants need graph memory not more tokens — 10:45am-11:05am · Graphs
- 2026 07 01 everett berry gtm engineering the technical bits — 10:45am-11:05am · AI in GTM
- 2026 07 01 chaitanya asawa from ambient documentation to clinical intelligence — 10:45am-11:05am · AI in Healthcare
- 2026 07 01 serena ge deepswe expert code datasets — 10:45am-11:05am · Agentic Engineering
- 2026 07 01 nishant gupta operating distributed inference systems at scale — 10:45am-11:05am · Inference
- 2026 07 01 pamela fox diagnosing agent failures in production — 10:45am-11:05am · Track M
- 2026 07 01 jennifer lee building safe payment infrastructure for machine to machine commerce — 10:45am-11:05am · Agentic Commerce
- 2026 07 01 safia abdalla the agent behind the curtain building the oz cloud agent platform — 10:45am-11:05am · AI Architects: AI Factories
- 2026 07 01 wallon walusayi ai engineering and governance 2026 trends — 10:45am-11:05am · Expo Stage 1
- 2026 07 01 willem pienaar your agent can t tell if it s right — 10:45am-11:05am · Expo Stage 2 NW
- 2026 07 01 ryan cooke no that s not a software factory — 10:45am-11:05am · Expo Stage 3
- 2026 07 01 session vector isn t enough hybrid search and retrieval for ai engineers — 10:45am-11:05am · Expo Stage 4 SE
- 2026 07 01 howie liu 100 000 aie startup battlefield presented by hyperagent — 10:45am-1:00pm · Startup Battlefield
- 2026 07 01 martin harrysson the agentic product development organization — 11:00am-12:00pm · CTO Circle
- 2026 07 01 nikita kothari mcps clis and skills choosing the right tooling layer for agentic development — 11:10am-11:30am · Agentic Engineering
- 2026 07 01 james russo html is all agents need — 11:10am-11:30am · Generative Media
- 2026 07 01 harshal bhangale why your ai agent needs a wallet agentic commerce on arc with usdc and nanopayments — 11:10am-11:30am · Agentic Commerce
- 2026 07 01 udi menkes why off the shelf ai doesn t understand money — 11:10am-11:30am · AI in Finance
- 2026 07 01 nader khalil state of the union why local why now 11 10am 11 30am track 4 420 — 11:10am-11:30am · Local AI
- 2026 07 01 yohei nakajima active graph agent runtime babyagi 4 — 11:10am-11:30am · Graphs
- 2026 07 01 aliisa rosenthal reverse engineering the ai buyer — 11:10am-11:30am · AI in GTM
- 2026 07 01 rashi agrawal guardrails first engineering member facing health ai — 11:10am-11:30am · AI in Healthcare
- 2026 07 01 frank coyle anthropic s cca exam as a field guide for agentic engineering — 11:10am-11:30am · Agentic Engineering
- 2026 07 01 qianru lao routing llm inference in production from engine signals to policy — 11:10am-11:30am · Inference
- 2026 07 01 chang liu tracing and debugging agents across systems with opentelemetry — 11:10am-11:30am · Track M
- 2026 07 01 dmitry buykin tribal dungeons of global shipping ai agents at global scale — 11:10am-11:30am · AI-Native Enterprises
- 2026 07 01 tisha chawla finops for ai agents who spent all the tokens — 11:10am-11:30am · AI Architects: AI Factories
- 2026 07 01 brandon waselnuk beyond rag see a relational context engine reduce token burn — 11:10am-11:30am · Expo Stage 1 NE
- 2026 07 01 zubin aysola aria how we built autoresearch with autoresearch — 11:10am-11:30am · Expo Stage 2 NW
- 2026 07 01 michael patterson the lethal trifecta is already on your developers laptops — 11:10am-11:30am · Expo Stage 3 SW
- 2026 07 01 matt gibiec your ai agent has no nervous system — 11:10am-11:30am · Expo Stage 4 SE
- 2026 07 01 michael grinich auth for agents unblock autonomous ai with auth md — 11:40am-12:00pm · Agentic Engineering
- 2026 07 01 ekaterina deyneka building an agentic video editor for mass consumer — 11:40am-12:00pm · Generative Media
- 2026 07 01 anil nadiminti when ai agents pay and sellers monetize building x402 apps for agentic commerce on aws — 11:40am-12:00pm · Agentic Commerce
- 2026 07 01 divakar kumar let s integrate ai agents in event sourced systems — 11:40am-12:00pm · AI in Finance
- 2026 07 01 ahmad osman demo glm 5 2 on dgx station frontier intelligence under your desk — 11:40am-12:00pm · Local AI
- 2026 07 01 mike phipps your moat is your data model — 11:40am-12:00pm · Graphs
- 2026 07 01 flora liu ai in gtm at notion — 11:40am-12:00pm · AI in GTM
- 2026 07 01 jared joselowitz shipping ai to a million patients without an a b test — 11:40am-12:00pm · AI in Healthcare
- 2026 07 01 anirban chatterjee guide verify solve the engineering discipline agentic development demands — 11:40am-12:00pm · Agentic Engineering
- 2026 07 01 ashok chandrasekar are llm performance benchmarks reliable — 11:40am-12:00pm · Inference
- 2026 07 01 ross wollman benchmarking vs code with vsc bench how to measure agent performance — 11:40am-12:00pm · Track M
- 2026 07 01 daniel kim all the things we have to do to satisfy your insatiable need for tokens — 11:40am-12:00pm · Inference
- 2026 07 01 khaled alashmouny what if your chip design team moved like a single body — 11:40am-12:00pm · AI Architects: AI Factories
- 2026 07 01 miguel gonz lez fern ndez the art of building verifiers for computer use agents — 11:40am-12:00pm · Expo Stage 1 NE
- 2026 07 01 sheilah kirui seeing the plumbing profiling vllm speculative decoding on nvidia blackwell — 11:40am-12:00pm · Expo Stage 2 NW
- 2026 07 01 kwindla kramer voice is the universal interface — 11:40am-12:00pm · Expo Stage 3 SW
- 2026 07 01 jake broekhuizen the next run should be better — 11:40am-12:00pm · Expo Stage 4 SE
- 2026 07 01 mike chambers harness engineering building the production cage for powerful domain agents — 12:05pm-12:25pm · Harness Engineering
- 2026 07 01 arturo nunez the next game engine won t have a manual — 12:05pm-12:25pm · Generative Media
- 2026 07 01 jan curn x402 isn t good yet — 12:05pm-12:25pm · Agentic Commerce
- 2026 07 01 vinoo ganesh how kepler built verifiable ai for financial services — 12:05pm-12:25pm · AI in Finance
- 2026 07 01 session local ai demos — 12:05pm-12:25pm · Local AI
- 2026 07 01 omri bruchim from systems of record to systems of context — 12:05pm-12:25pm · Graphs
- 2026 07 01 arman vaziri the building blocks of gtm orchestration — 12:05pm-12:25pm · AI in GTM
- 2026 07 01 vivek muppalla 200 million patient interactions later what the generic voice stack misses — 12:05pm-12:25pm · AI in Healthcare
- 2026 07 01 denys linkov benchmarking coding agents on new vs legacy code bases — 12:05pm-12:25pm · Agentic Engineering
- 2026 07 01 rita zhang vertical mobility building an ai inference platform that scales from mvp to trillion parameter workloads — 12:05pm-12:25pm · Inference
- 2026 07 01 tina manghnani design multi agent systems that actually work — 12:05pm-12:25pm · Track M
- 2026 07 01 pranay bhatia stop model shopping why ownership beats choice in the agent stack — 12:05pm-12:25pm · Inference
- 2026 07 01 archana kamath preferences benchmarks model routing for how teams actually build — 12:05pm-12:25pm · AI Architects: AI Factories
- 2026 07 01 giedrius steimantas the missing layer in agentic ai — 12:05pm-12:25pm · Expo Stage 1 NE
- 2026 07 01 ali adl tabatabai while you were generating the verification gap nobody talked about — 12:05pm-12:25pm · Expo Stage 2 NW
- 2026 07 01 ben dicken move fast and don t break things — 12:05pm-12:25pm · Expo Stage 3 SW
- 2026 07 01 sandhya subramani agents that forge their own tools self modifying ai in the wild — 12:05pm-12:25pm · Expo Stage 4 SE
- 2026 07 01 swyx latent space live the inference inflection from first principles — 12:30pm-1:30pm · Expo Stage 2
- 2026 07 01 brendan rappazzo loophole adversarial agents to stress test your morality — 1:30pm-1:50pm · Harness Engineering
- 2026 07 01 todd fisher while my guitar gently speaks — 1:30pm-1:50pm · Generative Media
- 2026 07 01 rodrigo coelho agent spending without controls the missing infrastructure layer for ai pa — 1:30pm-1:50pm · Agentic Commerce
- 2026 07 01 shawn chan build for the memo not the demo notes from 200 investment committees — 1:30pm-1:50pm · AI in Finance
- 2026 07 01 carter abdallah local models trust control optimization — 1:30pm-1:50pm · Local AI
- 2026 07 01 ritvik pandya ai learned execution graphs for real time anomaly detection and drift classification in apis — 1:30pm-1:50pm · Graphs
- 2026 07 01 alex bauer how juries and librarians can solve gtm s ai trust problem — 1:30pm-1:50pm · AI in GTM
- 2026 07 01 clay cockrell al is becoming the world s largest relationship therapist we can t afford to get it wrong — 1:30pm-1:50pm · AI in Healthcare
- 2026 07 01 dominik kundel codex behind the harness — 1:30pm-1:50pm · Agentic Engineering
- 2026 07 01 philip kiely what s new in inference engineering — 1:30pm-1:50pm · Inference
- 2026 07 01 chang liu evaluating and optimizing ai agents from observability to continuous improvement — 1:30pm-1:50pm · Track M
- 2026 07 01 josh leavitt from zero to ai native scaling ai across the org — 1:30pm-1:50pm · AI-Native Enterprises
- 2026 07 01 patrick debois coding agents don t scale themselves neither do your teams the rise of agent enablement — 1:30pm-1:50pm · AI Architects: AI Factories
- 2026 07 01 michael liendo trust but verify human in the loop for agents that actually matter — 1:30pm-1:50pm · Expo Stage 1 NE
- 2026 07 01 rowan christmas yolo mode safely microvm sandboxes for any agent — 1:30pm-1:50pm · Expo Stage 2 NW
- 2026 07 01 joshua mo your model is private your system isn t — 1:30pm-1:50pm · Expo Stage 3 SW
- 2026 07 01 rafael levi video discovery for agentic world model training — 1:30pm-1:50pm · Expo Stage 4 SE
- 2026 07 01 giselle van dongen every step you take every call you make the reliable agent stack — 1:55pm-2:15pm · Harness Engineering
- 2026 07 01 lina colucci voice agents with realtime video — 1:55pm-2:15pm · Generative Media
- 2026 07 01 anna spysz teaching agents to pay — 1:55pm-2:15pm · Agentic Commerce
- 2026 07 01 lucas palma we vetted 2 000 ai skills before they reached developers — 1:55pm-2:15pm · AI in Finance
- 2026 07 01 carter abdallah local models trust control optimization 1 55pm 2 15pm track 4 489 — 1:55pm-2:15pm · Local AI
- 2026 07 01 frank coyle why agentic systems need ontologies — 1:55pm-2:15pm · Graphs
- 2026 07 01 christopher burns how we got llms to recommend our open source library without paying or plug ins — 1:55pm-2:15pm · AI in GTM
- 2026 07 01 vasant kearney healthcare s agent bytecode x12 as the harness for ai agents — 1:55pm-2:15pm · AI in Healthcare
- 2026 07 01 arjun singh multiplayer agentic engineering enabling your whole team and your best agents to work together — 1:55pm-2:15pm · Agentic Engineering
- 2026 07 01 daniel svonava large clusters for small models — 1:55pm-2:15pm · Inference
- 2026 07 01 arun sekhar blast radius zero one command openclaw sandboxes in the cloud — 1:55pm-2:15pm · Track M
- 2026 07 01 brian lewis which ai startups actually land enterprise contracts lessons from evaluating 100 ai startups at millennium management — 1:55pm-2:15pm · AI-Native Enterprises
- 2026 07 01 r mi louf agent frameworks considered harmful — 1:55pm-2:15pm · Harness Engineering
- 2026 07 01 jan curn mcp doesn t suck your agent does — 1:55pm-2:15pm · Expo Stage 2 NW
- 2026 07 01 george he everyone talks about document search but what about results — 1:55pm-2:15pm · Expo Stage 4 SE
- 2026 07 01 sarah sanders we let an ai agent execute bash and lived to talk about it — 2:25pm-2:45pm · Harness Engineering
- 2026 07 01 keegan mccallum generative video at the speed of light — 2:25pm-2:45pm · Generative Media
- 2026 07 01 ahnaf prio the agentic commerce stack — 2:25pm-2:45pm · Agentic Commerce
- 2026 07 01 ramana siddanth emani your finance agent s bottleneck is you — 2:25pm-2:45pm · AI in Finance
- 2026 07 01 chris alexiuk compression at the edge — 2:25pm-2:45pm · Local AI
- 2026 07 01 james le video has no memory here s how we built one — 2:25pm-2:45pm · Graphs
- 2026 07 01 jeffrey wang knowledge systems the new gtm stack — 2:25pm-2:45pm · AI in GTM
- 2026 07 01 ayush bhardwaj trading desks to clinical trials parallels in applied vertical ai — 2:25pm-2:45pm · AI in Healthcare
- 2026 07 01 justin smith always on agents run production without the on call tax — 2:25pm-2:45pm · Agentic Engineering
- 2026 07 01 byung gon gon chun the frontier ai inference cloud for agents — 2:25pm-2:45pm · Inference
- 2026 07 01 ashu joshi operate agents safely at scale with enterprise governance — 2:25pm-2:45pm · Track M