Kairosys
LangGraph agentic orchestration platform with grounded RAG
Hybrid retrieval, reranking over PostgreSQL and pgvector, permissioned tool governance, structured outputs, citations, LangGraph, FastAPI, and Azure OpenAI.
For Hiring Teams
Senior AI engineer for teams building real production systems.
If you are hiring for a senior, staff, or principal-level AI engineering role, this page is the short version. My strongest fit is hands-on work across RAG systems, LangGraph-based orchestration platforms, real-time voice agents, eval infrastructure, and production hardening.
Senior AI Engineer
Staff / Principal level
RAG and LangGraph
Voice and eval systems
Southern California, remote. Hands-on builder with architecture depth and team leadership experience.
What I Build
Production AI systems with real operating constraints.
The strongest signal is the work itself: retrieval pipelines, orchestration layers, guarded voice systems, and evaluation loops that hold up outside of demo conditions.
Evidence QA
Evidence-bound document QA for legal and compliance use cases
Six-stage pipeline with injection detection, hybrid retrieval, confidence gating, verification, citation validation, and telemetry across PostgreSQL and Langfuse.
Resume Highlights
The short version hiring managers usually want first.
Hands-on daily in modern AI tooling, but with enough delivery and leadership depth to reason about production systems end to end.
Technical depth
- hybrid retrieval, reranking, embedding strategy, pgvector, Azure AI Search
- LangGraph, LangChain, LlamaIndex, structured outputs, MCP, model benchmarking
- real-time voice systems with LiveKit, Deepgram, Cartesia, ElevenLabs, asyncio concurrency
- golden sets, adversarial testing, regression suites, drift tracking, LangSmith, Langfuse, OpenTelemetry
Delivery history
- enterprise conversational AI deployment for an approximately 1,000-person insurer
- AI engineering lead and CTO work across product lines, retrieval systems, and production standards
- managed engineering organizations up to 16 directly and larger cross-functional groups in prior leadership roles
- 15 years of software engineering and platform leadership before the recent AI-native focus
How I Work
Hands-on in the codebase, comfortable with ownership.
I am strongest when the role expects both implementation and judgment: building the system, pressure-testing the architecture, and helping a team avoid predictable mistakes in retrieval, evals, releases, and observability.
Builder
Deep in architecture and implementation
FastAPI, Python, orchestration, evals, telemetry, provider integration, and production hardening are all normal daily work.
Operator
Understands what breaks after launch
Bias toward fail-closed controls, regression discipline, fallback logic, and instrumentation rather than demo-only optimism.
Teammate
Can lead, but does not need management theater
I have led teams, but the best current fit is still hands-on senior IC or IC-plus work where technical depth matters every week.
Hiring Contact
If you are hiring, send the role or book a short call.
For strong fits, the fastest path is usually the job description, team context, and the kind of system you need built or stabilized.