BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//smolboard//Event Schedule//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:AI Engineer Code Summit
BEGIN:VEVENT
UID:000000000000000018cd30f7817c42ce00000087@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20231111T140000Z
DTEND:20231111T144000Z
SUMMARY:Taming 40-Minute CI: Incremental Builds at Monorepo Scale
DESCRIPTION:Our monorepo CI took 40 minutes on a good day. This talk walks 
 through how we cut it to 6 minutes with content-addressed caching\, remote
  execution\, and a test-selection model — including the two migrations t
 hat failed first. You'll leave with a decision framework for which increme
 ntal-build investments pay off at which repo sizes\, and the graphs to con
 vince your platform team.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018ca4565d5d2cdff00000011@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261111T160000Z
DTEND:20261111T164500Z
SUMMARY:Opening keynote: Agents that ship
DESCRIPTION:What separates demos from dependable agent products. The gap is
  rarely the model. It is evaluation\, error handling\, and the unglamorous
  operational work that turns a promising prototype into something a team c
 an put in front of paying customers. Drawing on two years of shipping agen
 ts into production\, this keynote lays out what to build first and what ca
 n safely wait.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018c9f23f3b29217b00000025@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261111T170000Z
DTEND:20261111T172500Z
SUMMARY:Building reliable agent loops
DESCRIPTION:Most agent demos work once and fall apart on the second run. Th
 is talk walks through the loop itself: how to bound retries\, when to let 
 a tool failure propagate instead of swallowing it\, and how to keep the wo
 rking state small enough that a model can still reason about it. Includes 
 failure traces from three production agents and what each one taught us ab
 out where loops actually break.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018c9f23f56cd017700000028@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261111T170000Z
DTEND:20261111T183000Z
SUMMARY:Hands-on: malleable software with agents
DESCRIPTION:Software that users can reshape at runtime asks different quest
 ions of the architecture. In this hands-on session you build a small app w
 hose behavior an agent can rewrite\, then break it in the ways real users 
 will. Expect to spend most of the time in code rather than slides. Bring a
  laptop with Node 20 or later installed.
LOCATION:Workshop Room\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018c9f23f44b5a3f900000026@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261111T173000Z
DTEND:20261111T175500Z
SUMMARY:Evals that catch regressions before users do
DESCRIPTION:An eval suite that only runs before launch tells you nothing ab
 out the model you shipped. We cover building a graded set from real traffi
 c\, choosing metrics that move when quality moves\, and wiring the whole t
 hing into the deploy so a bad change never reaches a user. Bring a product
  with an eval gap and leave with a plan for closing it.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018c9f23f4de3bfe700000027@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261111T183000Z
DTEND:20261111T185500Z
SUMMARY:Substituting models safely
DESCRIPTION:Swapping the model underneath a live product is a migration\, n
 ot a config change. This session covers shadow traffic\, per-prompt scorec
 ards\, and the rollback path you need in place before the swap\, plus the 
 class of regressions that only show up under real user phrasing. We walk t
 hrough two swaps that went fine and one that had to be reverted at 2am.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018cb41a867f7c15100000000@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261111T183000Z
DTEND:20261111T193000Z
SUMMARY:Break
LOCATION:Workshop Room\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018ca4565e4edba2500000013@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261111T190000Z
DTEND:20261111T193000Z
SUMMARY:CI for your prompts
DESCRIPTION:How a team wires evals into every deploy. Each prompt change op
 ens a pull request\, runs against a graded set\, and reports a score diff 
 a reviewer can actually read. The talk covers how that set was built\, how
  it is kept current as the product changes\, and what the team does when a
  score drops but the change is still right.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018ca4565f2f96d0300000015@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261111T194500Z
DTEND:20261111T201500Z
SUMMARY:Grounding copilots without a vector db meltdown
DESCRIPTION:A field guide to retrieval that survives production traffic. Co
 vers chunking that respects document structure\, hybrid search for the que
 ries embeddings alone keep missing\, and the caching layer that kept p99 f
 lat while the corpus grew ten times. Includes the cost model that made the
  case for a smaller index and a bigger reranker.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018ca456602328be700000017@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261111T201500Z
DTEND:20261111T211500Z
SUMMARY:Lunch break
DESCRIPTION:Lunch is served in the main hall. Vegetarian\, vegan\, and glut
 en-free options are labeled at each station.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018ca45661050ec4b00000019@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261111T211500Z
DTEND:20261111T214500Z
SUMMARY:Latency budgets for multimodal agents
DESCRIPTION:Streaming\, speculative tool calls\, and where the milliseconds
  go. A multimodal agent spends its time in places a text-only one does not
 \, and the usual profiling advice misses most of them. This talk gives you
  a budget you can hold each component to\, and shows the three changes tha
 t took one agent from four seconds to under one.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018ca45661eea9fd10000001b@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261112T163000Z
DTEND:20261112T180000Z
SUMMARY:Workshop: build a coding agent fleet
DESCRIPTION:A hands-on session. You build a small fleet of coding agents th
 at split a task\, work in isolated worktrees\, and merge without stepping 
 on each other. We cover task decomposition\, the review step that catches 
 a bad agent before its work lands\, and how to keep the whole thing debugg
 able when it goes wrong. Bring a laptop with git and Node installed.
LOCATION:Workshop Room\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018ca45662db48bb70000001d@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261112T181500Z
DTEND:20261112T184500Z
SUMMARY:Regression detection at model-update speed
DESCRIPTION:Catching silent quality drops within hours of a model swap. The
  talk covers the sampling strategy that makes a small graded set represent
 ative\, an alerting threshold that does not fire on noise\, and the triage
  path once it does. Includes a walkthrough of a regression found\, diagnos
 ed\, and reverted in a single afternoon.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018ca45663cb495df0000001f@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261112T190000Z
DTEND:20261112T193000Z
SUMMARY:Serving embeddings for 40k QPS
DESCRIPTION:War stories from a production platform team. Reaching 40\,000 q
 ueries per second meant giving up on a single index\, batching aggressivel
 y at the edge\, and accepting a staleness window the product team had to s
 ign off on. This talk covers the architecture\, the two rewrites it took t
 o get there\, and the monitoring that keeps it honest.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018ca809ebe92314b00000000@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261112T203000Z
DTEND:20261112T210000Z
SUMMARY:Live-coding a conference copilot
DESCRIPTION:Building an event copilot live on stage with tool calls against
  a real schedule. Starting from an empty file\, we wire a model to the con
 ference API\, give it the four tools it needs\, and watch it answer attend
 ee questions in real time. Nothing is pre-recorded\, so expect at least on
 e thing to break and get fixed in front of you.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
BEGIN:VEVENT
UID:000000000000000018ca45664b004e3b00000021@smolboard
DTSTAMP:20260828T040829Z
DTSTART:20261112T213000Z
DTEND:20261112T221500Z
SUMMARY:Closing panel: prototype to production
DESCRIPTION:A moderated conversation with speakers from across the program.
  The panel takes questions on the hardest part of the last two years: deci
 ding when a prototype is ready\, what to cut to get it there\, and how a t
 eam stays confident once the model underneath it keeps changing. Submit qu
 estions during the day and the best ones go to the panel.
LOCATION:Main Stage\, San Francisco\, CA
URL:https://www.smolboard.app/ai-engineer/ai-engineer-sandbox
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR