Ali Khatami
Technical Product Leader · Staff-Level Platform Engineer
Technical Product Leader with Staff-level engineering depth. Six years building developer platforms, AI/LLM automation systems, and API governance tools at enterprise scale. Rebuilt Chase's entire API onboarding lifecycle—cutting governance from 2 months → 1 week and documentation from 4–5 weeks → minutes using LLM pipelines, RAG, graph-based scoring, and OpenAPI-first architecture. Founder of nfrax, a 100+ star open-source SDK ecosystem powering backend, AI, and fintech workflows.
Governance approvals: ~2 months → ~1 week · Documentation: ~4 weeks → minutes · ~90% fewer governance/compliance errors
Professional Experience
VP / Lead Software Engineer + Technical Product Lead
JPMorgan Chase & Co.New York, NY- Designed the full automation architecture for API governance, documentation, and readiness scoring using LLMs, RAG, and graph-based state machines.
- Compressed governance from 2 months → 1 week via deterministic scoring, evidence capture, and automated remediation tied directly to OpenAPI updates.
- Built parallelized LLM documentation generation with hallucination mitigation, ML-based validation, and spec-driven consistency checks—cutting cycles from 4–5 weeks → minutes.
- Architected graph-driven progression models for governance states, readiness scoring, compliance gates, and remediation workflows.
- Built MCP-powered agentic systems answering ownership, readiness, and standards queries across thousands of APIs.
- Drove adoption across the firm; platform is now the CCB standard for onboarding and readiness.
- Collaborated with partner integration teams building Disney eGift APIs, Amazon Card APIs, and loyalty systems using Java, Spring Boot, AWS, Terraform, and resilient event-driven designs.
- Recognized in executive briefings and CCB townhalls; work presented to CTO Gill Haus.
Technical Product Manager — API Platform
JPMorgan Chase & Co.New York, NY- Mapped organizational friction across onboarding, governance, documentation, and partner readiness.
- Defined multi-phase modernization roadmap moving documentation from inconsistent Word docs → standardized Markdown → spec-driven autogeneration.
- Shifted OpenAPI specs to source-of-truth, improving governance accuracy and partner trust.
- Authored PRDs and NFRs covering auditability, SLOs, risk controls, and fail-safe behaviors.
- Designed API-driven documentation storage and retrieval, replacing file-passing with structured, scalable systems.
- Result: Platform became enterprise standard for API onboarding and partner delivery.
Senior Associate Engineer — ML & Platform
JPMorgan Chase & Co.New York, NY- Built ML-driven risk forecasting pipelines; migrated Pandas workloads → PySpark for scalability.
- Developed full-stack platform (Python/Django + React) for cloud cost visibility and resource planning, adopted org-wide.
- Created reusable CI/CD patterns used across CIB engineering teams.
Founder & Lead Developer
nfrax (Open Source · 100+ Stars)Remote- Built 4 MIT-licensed SDKs: svc-infra (backend framework), ai-infra (LLM/RAG workflows), fin-infra (fintech integrations: Plaid, Teller), robo-infra (robotics/IoT toolkit).
- Designed type-safe interfaces, retries, caching, service abstraction layers, and resilience patterns.
- Unified APIs that reduce boilerplate and speed up delivery; adopted by open-source developers and early startups.
Founder & Lead Engineer
QuantTech MarketsNew York, NY- Built algorithmic trading engine using Python, ML, and NLP models—outperformed the S&P 500 by 5–7%.
- Sold the backend to a trading firm while retaining usage rights.
- Built full-stack platform (React + Python) with reproducible backtesting, portfolio analytics, and automated deployments.
Key Projects
AutoChannel
API governance + documentation automation — ~80% faster governance reviews and consistent OpenAPI readiness across 4,000+ APIs.
LangGraph, MCP, LLMs, Vector Search, OpenAPI
nfrax
SDKs that remove backend boilerplate — 3 MIT-licensed SDKs that cut boilerplate and speed delivery.
Python, FastAPI, MCP, Vector Search, Plaid
MCP Servers
Standardized tool invocation for agents — Repeatable, secure agent workflows adopted org-wide.
MCP, TypeScript, Python, VS Code API