TANAY.SHAH
● PRODUCTION · 2026 · FOUNDING-ENGINEER WORK · UPDATED 2026-05-10

Structured AI · Founding Engineer.

Fourteen weeks of agent infrastructure, sandboxed runtimes, an iOS app, and durable chat — all in production.

// 01 — WHY I BUILT IT
THE PROBLEM

Joined at the founding-engineering stage with a clear mandate: ship the agent infrastructure, the iOS client, and the deploy pipeline that production needed. Most of the engineering decisions had compound constraints — enterprise compliance, cross-platform parity, deploy speed — that rewarded engineers who could move fast across the stack.

// 02 — THE APPROACH
THE WORK

Fourteen weeks of velocity: 1,035 commits, ~74/week median, peak week of 145. The work spanned backend (FastAPI + Postgres + Redis + gRPC bridge), agent loop (LangGraph + Claude with on-demand tool loading), code execution (in-house sandbox built for Zero Data Retention compliance), iOS / iPadOS client from zero, and CI work cutting staging deploys to under seven minutes.

Most architectural detail beyond the public timeline is NDA-bound. The deeper version is available in 1:1 conversation under appropriate scope.

// 03 — KEY DECISIONS
WHAT I CHOSE & WHY
DECISION · 01

In-house sandbox over hosted PTC

Anthropic's hosted Programmatic Tool Calling doesn't support Zero Data Retention, and enterprise clients required ZDR. Built a request-interception layer that handles PTC traffic locally — untrusted code runs in a hardened in-house pod (JSON-RPC over child subprocess, AST validation, bubblewrap + seccomp); customer data wipes immediately and never reaches the model provider.

DECISION · 02

Multi-vendor agent (Claude reasoning + Gemini vision)

Anthropic Claude drives the reasoning loop; Google Gemini Vision is delegated for high-resolution document inspection through a multi-tile rendering pipeline that yields significantly higher effective resolution than naive full-page rendering. Different providers, different strengths — the agent uses each where it wins.

DECISION · 03

Synthesis over framework adoption

Built three prior agents on LangGraph Deep Agents, raw LangGraph nodes, and Anthropic's Claude Code SDK loop. None hit the bar. The shipped agent is a synthesis: takes the strongest pattern from each (prefetch-backed tools and server-side compaction, workspace + execute model, minimal-tool philosophy, adaptive thinking) and strips out every line of framework overhead.

// 05 — STATE OF THE ART
2026 BLEEDING-EDGE TECH
LangGraph + Anthropic Claude (Opus / Sonnet)

Modular agent loop with on-demand tool loading and persistent skill / memory injection. Cache-friendly via Anthropic's ephemeral cache_control directive — the right primitive for cost-sensitive long-context agents.

Multi-tile vision delegation (Gemini Vision)

Construction drawings are 36×24-inch sheets — naive full-page rendering loses critical detail. Multi-tile rendering yields significantly higher effective resolution at the same per-tile cost. Multi-vendor design where each provider plays its strength.

Hardened Python sandbox (bubblewrap + seccomp + AST validation)

Linux user-namespace isolation + syscall filtering + pre-execution AST static analysis. Same-class hardening as production code-execution platforms (Replit / E2B), built in-house for ZDR-compliant enterprise deployment.

Programmatic-Tool-Calling request interception

Anthropic's hosted PTC doesn't support Zero Data Retention. Custom interception layer keeps execution local while preserving Claude's PTC code-quality. ZDR compliance without sacrificing model capability.

gRPC + Protobuf bridge between API and agent pod

Cross-process typed RPC at scale. Protobuf schemas double as the API contract — no drift between client and server.

Event-sourced durable chat (session_events + replay cursor + DLQ)

Chat survives network blips, pod restarts, and deploys without lost messages. Replay-on-reconnect is the modern primitive for production conversational UI.

Azure Container Apps Jobs + ACR Tasks (sub-7-min deploys)

Container-native serverless compute for the agent pod. ACR Tasks build with Blacksmith BuildKit cache. CPU-only torch + blobless wipe-guard clone cut staging deploys from 10+ min to under 7.

// 06 — MEASURED
NUMBERS THAT MATTER
Commits / 14 wks
1,035
Author-filtered, single email
Lines authored
~832K
At HEAD
PRs merged
30
GitHub
Repo insertions
~37%
Of all-author insertions ever
// 07 — IF I DID IT AGAIN
LESSONS · WHAT I'D CHANGE
  • Founding-engineering velocity isn't typing speed; it's the ability to make 'good enough now, refactor next sprint' calls without stopping the line. Every week I shipped, I also paid down a known piece of debt I'd taken the prior week. Net: forward motion every Monday.
  • Multi-vendor agent design is increasingly important — single-provider lock-in is a liability for both compliance and capability. The pattern of 'reasoning model + delegated vision model' will keep generalizing.
  • The most valuable infrastructure I shipped was the CI work. Cutting staging deploys from 10+ minutes to under 7 changed how the whole team iterated.
// 08 — STACK
THE TOOLS
BACKEND
Python · FastAPI
STORE
PostgreSQL · asyncpg · pgvectorRedis
RUNTIME
Docker · multi-stage
INFRA
Azure (Container Apps · ACR)
AGENT
LangGraph · Anthropic ClaudeGoogle Gemini Vision
TRANSPORT
gRPC + Protobuf
SECURITY
bubblewrap · seccomp · AST validation
WEB
SvelteKit + Svelte 5
IOS
Swift / SwiftUI