1/5
No QA test result, issue report, video, or screenshot was provided. The brief introduction did not document any testing or address the required deliverables.
GitHub QA: #2824 @ 58ba5edfdae2 · Aug 5, 2026
AI Engineer | Production RAG, LLM Agents & Evaluation Systems
first conversation is free, sign up to message Melvin
I am an AI Engineer focused on building reliable production LLM systems in Python, TypeScript, and Rust. I develop retrieval-augmented generation pipelines, LLM agents, contextual automation, evaluation frameworks, and secure tool integrations. In my current role, I built an LLM-driven Actions layer that recommends SaaS connectors and generates contextual one-click actions across 10+ integrations. I also implemented a Qdrant-based RAG pipeline over 20,000+ records, reducing end-to-end cold-start latency from 45 seconds to under 7 seconds. I have delivered 35+ AWS Lambda functions and designed an OAuth-based cloud integration layer supporting multiple providers. I also build tools for AI reliability and observability, including agentic RAG scorecards, prompt-injection red-team harnesses, OpenTelemetry tracing, retrieval evaluation, and cost/latency controls. My work includes Rust-based AI agent trace analysis and regression localization. I can help with LLM application development, RAG architecture, agent tooling, API integrations, AI evaluation, and production reliability.
Software Developer at Xavier AI (Feb 2025 - Present)
No QA test result, issue report, video, or screenshot was provided. The brief introduction did not document any testing or address the required deliverables.
GitHub QA: #2824 @ 58ba5edfdae2 · Aug 5, 2026