Trinath Reddy · Senior Software Engineer II · AI/ML
I build intelligent systems for the AI era.
AI Engineer · GenAI · Agentic AI · RAG · Cloud · Production Systems
How a request becomes intelligence.
A production AI request passes through identity, policy, routing, context, inference and evaluation — with cost, safety and observability running across every layer.
Example request · simulated, not a live AI service
“Analyze this quarter’s sales and explain why revenue dropped.”
Business analysis · medium complexity · balanced model
—
Input 2.4K · retrieved 1.6K · output 0.8K
—
Input $0.006 · output $0.009 · tools $0.003
—
Illustrative end-to-end request latency
—
Groundedness 94 · relevance 91 · safety 99
Illustrative budgets, scores and costs · animation time is not inference latency.
From ideas to intelligent systems.
Good AI isn't just about models.
Prompting gets a demo. Engineering gets a system people can trust on a Monday morning.
From notebook → production.
I specialise in the unglamorous 90%: evals, latency budgets, GPU bills, tracing and the rollback plan.
An evolving system, not a résumé.
Software Engineer
Foundations: APIs, distributed systems, shipping discipline.
AI/ML Engineer
Models meet data pipelines and real users.
GenAI Engineer
LLMs, prompting, retrieval and the first agents.
AI Systems Architect
Designing multi-agent, RAG and eval systems end-to-end.
AI Platform Architect
Platforms that let whole teams ship AI safely.
The stack, connected.
07 — Build with me