Consulting · Builds · Team training
Senior MLOps and AIOps engineer with 8+ years in production systems. I have shipped multi-agent incident triage, RAG support bots for 80,000+ users and fraud detection pipelines. Now I help companies do the same.
Support, ops, sales or internal tools teams
A working agent on your data: RAG over your docs, tool calls into your systems, human approval on risky actions, guardrails, tracing and cost limits.
You get: Deployed agent, evaluation set, runbook and handover
Founders and engineering leads planning an AI feature or product
We look at your idea, data and stack together, pick the right model and serving setup, and cost it out before you build.
You get: Written roadmap, architecture diagram, build-vs-buy call, cost estimate
Teams with models stuck in notebooks or running without monitoring
CI/CD for models, model registry, gated promotion, drift monitoring and retraining, GPU serving on Kubernetes, observability.
You get: Pipelines in your cloud (AWS, Azure or GCP) with Terraform and docs
Teams whose GPU, LLM API or cloud bill is growing faster than usage
Model routing between frontier and open-weight models, GPU right-sizing, leftover resource cleanup, inference cost guardrails.
You get: Findings report with savings estimate and a fix plan
Engineering teams moving into MLOps, LLMOps or agents
Live, hands-on program built on your stack and use cases, from fundamentals to production labs.
You get: Custom syllabus, live sessions, labs, recordings, assessments
Client delivery · US database company
Multi-agent system that reads production alerts, pulls the right runbooks, suggests a fix and waits for human approval before acting.
Client delivery · FinTech company
Production chatbot answering KYC, billing and policy questions for 80,000+ active users, handing over to a human when it is not sure.
Client delivery · UK payments processor
Champion vs challenger MLOps platform for fraud detection, forecasting and customer lifetime value.
Tell me what you want built or fixed. Two minutes.
30 minutes to understand your data, stack and goal. Free.
Scope, architecture, timeline and a fixed price or weekly rate.
Weekly demos, then deployment, docs and a handover session.
An AI agent for your team, a roadmap for your AI product, or help getting models into production.