Senior Full Stack & AI Engineer
10+ years building customer-facing products at scale, with React and Angular on the front end and Node.js, Java or Python behind them. Lately I've been adding real-time AI features: streaming assistants, voice, and live dashboards.
I've spent most of my career on high-traffic, customer-facing apps and the internal tools behind them: multi-step application flows, self-service portals, micro-frontend platforms and data-heavy dashboards. I care about the boring parts that make them good: fast pages, accessible UI, clear APIs and tests that catch things before customers do.
For the last couple of years most of my work has involved AI. That means connecting LLMs to real systems safely: retrieval over company documents, assistants that can take actions with the customer's confirmation, and the evaluation and guardrail work that gets a feature through security review. I've also done a lot of real-time work with WebSockets, Kafka and streaming responses, so AI features feel instant instead of bolted on.
I've led teams as the technical lead: setting coding standards, reviewing PRs, mentoring onshore and offshore developers, and owning CI/CD pipelines and production releases from build to sign-off.
A streaming chat assistant that answers from a document knowledge base with cited sources (vector search with pgvector), and uses schema-validated tool calls to act through existing REST APIs, with a human confirmation step before any write.
Live speech-to-text streamed over WebSockets into an LLM that suggests answers and surfaces relevant docs mid-conversation, then writes a summary at the end. Also prototyped a voice agent on WebRTC with barge-in handling and time-to-first-token tuning.
An LLM pipeline that turns unstructured PDFs into typed JSON using function calling and Pydantic validation, with automatic retries on schema errors and a human-in-the-loop review UI before anything is saved.
Kafka events fan out to the browser over Socket.IO, scaled horizontally with Redis Pub/Sub. A rolling window feeds an LLM that flags anomalies and writes plain-English summaries as new events arrive.
Let's talk.
I'm looking for senior full stack or AI engineering roles. If you're hiring or just want to compare notes on shipping AI features, I'd like to hear from you.