~/talk

Day 2

on YouTube 09h 28m
Talkat AI Engineer Europe 2026 · 2026-04-08..04-10· [large-language-models, ai-assisted-development]

speakers

name:Google DeepMind
links:none on file
name:OpenClaw
links:none on file
name:Adaption
links:none on file
name:Arize
links:none on file
name:Cloudflare
links:none on file
name:Kaspar Hidayat
links:none on file

resources

video:https://youtu.be/m12vGjfbNlo
slides:none on file

summary

Titles that promise specifics win: Arize shows Alex 2.0 claiming explicit planning tools and four states (pending / in progress / blocked / completed), large JSON abstractions, and reliable agent checkpoints; Cloudflare demos Code Mode TypeScript agents that write code instead of many tool calls and run in V8 isolates secure sandbox for fast, safe exec; OpenClaw maps plugined composable coding primitives; Google DeepMind frames deterministic boundaries for non‑deterministic models; Robot Company runs teleoperated robots + embodiment data and measures how much teleop cuts the autonomy gap.

AI-written summary · about

chapters (39)

  1. 0:00Introduction
  2. 7:01Opening remarks by Kaspar Hidayat
  3. 8:36SallyAnn DeLucia (Arize AI): Alyx planning states, large JSON abstractions, and reliable agent checkpoints
  4. 24:54Timothy Lin (Resaro): scenario specific evals, ODDs, and synthetic data for mission critical AI
  5. 35:50Abhishek Kankani (Cloudflare): Code Mode, one shot TypeScript agents, and secure V8 execution
  6. 48:50Tejas Kumar (IBM): agent harness primitives, loop control, and trust over black box models
  7. 1:07:54Break: Morning coffee
  8. 1:32:42JJ Geewax (Google DeepMind): applied AI at scale with deterministic boundaries around non deterministic models
  9. 1:53:33Geoff Huntley: everything is a factory, software abundance, and leaner organizations
  10. 2:11:46Vincent Koc (OpenClaw): OpenClaw foundations, plugins, and composable coding agent primitives
  11. 2:22:28Vishnu (Vish) Hari (Ego AI): AI characters with personality, memory, and a more human inner life
  12. 2:31:53Ben Guo (Zo Computer): personal cloud infrastructure, owned agents, and a software company for everyone
  13. 2:42:50Matthias Lubken (Tavon AI): embedding PI, simple agent loops, and OpenClaw style extensibility
  14. 2:56:27Josh Newton (Microsoft AI): design as the edge, AI as a magic pencil, and taste over slop
  15. 3:07:04Sam Bhagwat (Mastra): production agent patterns for customer, internal, and developer workflows
  16. 3:25:50Pierre-Loic Doulcet (LlamaIndex): LlamaParse failure modes, whitespace loops, and parsing at internet scale
  17. 3:38:09Jun Yu Tan (Tusk): Fence, OS level guardrails, and deterministic safety for coding agents
  18. 3:48:32Break: Lunch
  19. 5:01:45Sara Hooker (Adaption Labs): adaptive intelligence, dynamic data, and moving past brute force scaling
  20. 5:17:28Vincent Wu (MiniMax): agents that schedule their own compute and inference exchanges
  21. 5:29:44Daniel & Siddharth Krishnan (The Robot Company): teleoperated robots, embodiment data, and closing the autonomy gap
  22. 5:41:06Justin Baird (Tesseract) with Kai Ming: shared autonomy, BCI painting, and embodied creative agency
  23. 5:55:52Aravind (SK) Kandiah (Bifrost): sim generated worlds, robotics evals, and faster edge case discovery
  24. 6:10:11Julia Kim (OpenGraph Labs): sensorized humans, touch data, and training better humanoids
  25. 6:19:44Suveen Ellawela (Cortex AI): full stack robotics, data pipelines, and real world evaluation
  26. 6:29:42Break: Afternoon
  27. 6:30:15Keziah: guided sensory meditation with a vibe coded visualizer
  28. 6:57:36Jay Demetillo: prompt fatigue, human context, and design judgment AI cannot replace
  29. 7:09:54Alex Lee (Magic Patterns): AI native design systems, brand fidelity, and shippable UI generation
  30. 7:20:02Sabina Cabrera (MagicPath): balancing product and polish in code backed design workflows
  31. 7:34:01Priyaa Kalyanaraman (Lica World): design intelligence across taste, iteration, and layered editability
  32. 7:44:55Conor Brennan-Burke (Hyperspell): building a company brain from docs, Slack, and execution traces
  33. 7:56:35Heng Hong Lee (Lightsprint): plan, preview, orchestrate, three primitives for cloud coding agents
  34. 8:07:51Louis Knight-Webb (Vibe Kanban): planning and review as the human loop around parallel agents
  35. 8:19:42Harsha Khurdula (Interfaze AI): deterministic developer tasks with specialized encoders and LLM decoders
  36. 8:31:07Hrishi Olickel (Southbridge): high context agent runtimes, declarative budgets, and legacy system reliability
  37. 8:44:30Henry Mao (Smithery): MCP, CLIs, and the harness era of agent agency
  38. 8:56:30Rach Pradhan (Independent): reliable agentic workflows, code intelligence, and parallel agent systems
  39. 9:08:45Closing remarks by Agrim Singh

related talks

~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.