AI Radar
Simon WillisonNative is now the future of mobile at Shopify— Shopify is abandoning React Native for native Swift and Kotlin codebases, concluding that the cost of maintaining two platforms outweighs the benefits of shared code.arXiv cs.AIOpenDiscoveryTrace: Process Traces for Evaluating AI Scientist Workflows— OpenDiscoveryTrace releases 558 complete AI scientific agent trajectories to enable auditing of reasoning processes, which current benchmarks ignore by only evaluating final outputs.arXiv cs.AISubagents vs Agent Skills: Executing Reusable Knowledge for Long-Horizon Agentic Tasks— The paper compares subagents and agent skills for long-horizon tasks, analyzing how reusable knowledge packages are executed to improve agent performance on complex workflows.Zvi MowshowitzGPT-6 Astra: The System Card, Alignment and What Comes Next— OpenAI claims Astra is the most intelligent and aligned available model, a stance Zvi Mowshowitz scrutinizes for its alignment implications.Zvi MowshowitzAstra Is Hard to Monitor— Zvi Mowshowitz argues Astra is hard to monitor, challenging OpenAI’s central message regarding the model's safety and alignment status.Simon WillisonNative is now the future of mobile at Shopify— Shopify is abandoning React Native for native Swift and Kotlin codebases, concluding that the cost of maintaining two platforms outweighs the benefits of shared code.arXiv cs.AIOpenDiscoveryTrace: Process Traces for Evaluating AI Scientist Workflows— OpenDiscoveryTrace releases 558 complete AI scientific agent trajectories to enable auditing of reasoning processes, which current benchmarks ignore by only evaluating final outputs.arXiv cs.AISubagents vs Agent Skills: Executing Reusable Knowledge for Long-Horizon Agentic Tasks— The paper compares subagents and agent skills for long-horizon tasks, analyzing how reusable knowledge packages are executed to improve agent performance on complex workflows.Zvi MowshowitzGPT-6 Astra: The System Card, Alignment and What Comes Next— OpenAI claims Astra is the most intelligent and aligned available model, a stance Zvi Mowshowitz scrutinizes for its alignment implications.Zvi MowshowitzAstra Is Hard to Monitor— Zvi Mowshowitz argues Astra is hard to monitor, challenging OpenAI’s central message regarding the model's safety and alignment status.
Agentic AI · Systems Leadership

Harness engineering
for agents that act —
and answer for it.

Long Yang leads Intuit's agentic AI transformation. Nights and weekends, he has solo-built four multi-agent systems that run in production — trading, market intelligence, video and personal finance — plus the A2A mesh that connects them. 2,501 commits. 13,415 tests. Two of the operator consoles are open on this page, live, right now.

01 / About

Built for reliability.

Anyone can wire up an agent that demos well — the hard part is everything around the model. Long treats evals as the ship gate, context as a first-class engineering concern, and restraint as a capability: his systems earn trust by declining to act, degrading honestly, and logging every judgment.

Agentic AI
Harness engineering

One of the earliest engineers to put AI agents into production at Intuit — now leading that work end to end: MCP, multi-agent architectures, and looping engineering, where agents close their own feedback loop instead of being re-prompted by hand.

Systems
A decade in production

Kafka, Snowflake, FastAPI, pgvector — a decade of owning distributed systems end-to-end before "AI engineer" was a title. Agentic AI is that same discipline applied to a harder problem: reliability under non-determinism, at scale.

Leadership
Direction others build on

Leads a team of six to ten engineers at Intuit on agent design, evaluation harnesses and the orchestration layer beneath them. The patterns he set for orchestration, observability and crash recovery became the standard for the wider organisation's AI work, not just his own team's. Earlier, built and scaled a seven-engineer team at SVB.

02 / Projects

Four systems, one person.

Four systems. 2,501 commits and 13,415 tests — evenings and weekends, alongside the full-time role above. Each one engineered to survive a bad decision, not just make good ones.

Trader Joe
Autonomous Trading

A five-role agent loop — analyst, trader, challenger, evaluator, reviewer — paper-trading live market data, which this month measured its own alpha for the first time and spent a week fixing what that number exposed.

The hard part: hard risk limits enforced in code the model can't override, and a self-improving loop where a repeated mistake gets promoted from a logged lesson into a permanent, hard-coded gate.

Multi-agentRAGLLM evals
5Domain experts
Read the case study →
Family CFO
Autonomous Finance

A financial agent that gives a household real planning power over its own money — budgets, net worth, cash-flow forecasts, anomaly detection — running unattended against live data from any Plaid-supported bank.

The hard part: turning messy real-world transaction data into numbers a family can trust every day, with no one checking its work.

Function callingData integrityUnattended ops
6Specialist agents
Read the case study →
TMZ
Market Intelligence

Five replaceable market-intelligence experts — collection, enrichment, novelty, evaluation, signal — wrapping a production pipeline Trader Joe already reads from today, with a ReAct orchestrator built and staged behind them.

The hard part: re-architecting a live production pipeline in place — every expert ships first as a byte-identical wrapper, proven against a consistency test, before any orchestration logic is allowed to change what actually runs.

ReAct orchestratorA2A protocolExpert agents
5Market-intelligence experts
Read the case study →
Video Factory
Multi-Agent Production

38 modules, one GPU orchestrator, two production modes: fully autonomous for trending news — and for the deeply personal work, an AI that interviews a family, live, by voice, to fill in the gaps in their own story.

The hard part: an interview agent that has to ask a family the exact right question — grounded in real gaps in their own photo timeline, never an invented memory — plus a custom LLM router and MCP tool servers built from scratch to run the other 38 modules underneath it.

Agent orchestrationMCPGPU routing
26+Autonomous agents
Read the case study →
02b / Ecosystem

How they connect

Four independent systems, four separate reasons to exist. Wall Street is the protocol that lets them optionally exchange data and consult each other — a real contract between independently-owned systems, not just a hub node in a diagram, where every connection is opt-in and every system keeps working exactly as designed if its peers go dark.

Wall Street3 agentsFamily CFO6 agentsTraderJoe5 agentsTMZ5 agentsVideo Factory26+ agentsStandalone by design
How these systems collaborate
  • Mutual support, never mutual dependency: every system's core function works with all other peers offline.
  • Any peer being unreachable degrades to an explicit 'not connected' state — it never crashes or blocks the caller.
  • Cross-system calls are read-only by default; remote conclusions can only produce proposed actions gated by local user confirmation.
  • No shared databases across systems — HTTP contracts only, and the contract schema is the privacy whitelist.
  • Every capability is env-gated and off by default; off means byte-identical behavior to before integration.
Read the full design
03 / Experience

A decade in production.

The actual problem

The central engineering problem was not making it work — it was making it fail safely. A replay that starts and then dies leaves production misconfigured in a way no error message will ever report.

Agentic AI Engineer

IntuitNov 2025 – Present

Leads the transformation of Intuit's Go-to-Market data platform from traditional ETL to an AI-native, agentic architecture, with a team of six to ten engineers. Design owner and technical lead for a production agent that performs a dangerous data operation which used to take an engineer one to three days of babysitting — it runs unattended for days, survives its own server being destroyed and replaced mid-run, coordinates three teams over Slack and JIRA, writes to production through an audited approval process, and independently verifies the data landed correctly at every destination. Alongside it: a CI-gated eval pipeline that replaced days of manual QA with minutes of autonomous execution, and agentic tool chains that took marketing data onboarding from weeks to hours across pipelines carrying tens of millions of records at peak.

Read how it works

Senior Software Engineer

Capital OneJul 2023 – Jun 2025

Migrated legacy infrastructure to AWS CDK and CloudFormation. Built an AI-powered CI/CD pipeline using retrieval-augmented generation to cut pipeline-failure resolution time.

Senior Software Engineer

SVB Financial GroupJun 2020 – Jun 2023

Led design of mission-critical information-reporting systems for next-generation online banking. Built and managed a team of 7+ developers.

Senior Software Engineer

CiscoJan 2015 – Jun 2020

Designed Chassis-View, an industry-first graphics-oriented network device management system. Led the migration from a monolithic architecture to microservices.

Long has been engineering and shipping software in Silicon Valley since 2015 — infrastructure, data platforms, and now agentic AI. Outside of Intuit, the four systems above, and the protocol that connects them, he's quietly incubating a separate consumer venture, Game On — a different kind of bet, on its own clock.

He believes AI, built thoughtfully, will make the world meaningfully better — more productive, more creative, more human.

— Résumé & Contact

Let's talk.

Recruiter, investor, or fellow builder — the résumé is one click away. The assistant in the corner has read every write-up on this site, can check what the trading desk did today, and will draft the intro email for you.

LinkedIn ↗[email protected]