Charlie M.

AI & Backend Engineer | Tech Lead & Engineering Manager (100% Job Success, Top Rated Plus, Expert Vetted, 4,200+ Upwork hours). I design and deliver RAG pipelines, multi-agent LLM apps, FastAPI microservices, and MLOps systems on AWS/GCP/Azure—enterprise-ready, scalable, and secure. What I Build RAG & LLM Applications — OpenAI, Azure OpenAI, AWS Bedrock, Anthropic; LangChain/LlamaIndex; vector search and vector databases (Pinecone, Weaviate, FAISS); prompt engineering, guardrails, and evaluation harnesses for quality, latency, and cost. Multi-Agent Workflows — tool design, orchestrators, short/long-term memory, function calling, and policy controls for support, operations, and research automation. Backend Microservices & APIs — Python, FastAPI, REST/GraphQL, event-driven patterns with Kafka/SQS, API Gateway, auth, rate-limiting, and observability. MLOps — CI/CD for ML, MLflow & model registries, feature storage, batch/real-time inference, A/B and shadow deployments, monitoring & alerting. Cloud Architecture (AWS/GCP/Azure) — Docker, Terraform/IaC, AWS Lambda/ECS/S3/IAM/CloudWatch, cost controls, security baselines, secrets management. Leadership & Delivery Operate as IC, Tech Lead, or Engineering Manager. I handle system architecture, technical roadmaps, sprint planning, backlog/prioritization, stakeholder comms, code reviews & mentoring, estimation, and KPI-driven delivery (accuracy, latency, cost, reliability). Clear documentation (design docs, ADRs, runbooks) is included. Recent Wins (representative) Policy-safe multi-agent workflow for research/drafting → 100+ hours saved per month; measurable quality gates and auditability. Streaming pipeline modernization (event-driven) → data freshness improved from daily batches to near real-time; reduced ops toil. RAG knowledge system + eval harness → higher factual accuracy, lower support handle time; dashboards for latency and cost. Typical Engagements Production RAG MVP (OpenAI/Azure/Bedrock) — ingestion/chunking, retrieval, prompt strategy, evals, vector DB, metrics, and handoff. Multi-Agent Automation for Support/Ops — tool stack, orchestrator, safety/policy rules, runbooks, latency/cost dashboards. FastAPI Microservice + CI/CD — service template, tests, auth, Docker, GitHub Actions, deploy to AWS, logging/metrics. MLOps Starter (Databricks/MLflow) — experiment tracking, model registry, deployment patterns, sample pipeline. RAG/LLM Stack Audit — architecture review, eval gaps, reliability/cost plan, prioritized recommendations. Industries & Data SaaS, data platforms, operations/support tooling, analytics, internal knowledge bases, ETL/ELT modernization, research automation. Comfortable with PII handling patterns and least-privilege designs. Tooling & Stack Python, FastAPI, Docker, Terraform, AWS (Lambda/ECS/S3/IAM/CloudWatch), GCP, Azure, Kafka, Postgres, Redis, Databricks/MLflow, Pinecone, Weaviate, FAISS, LangChain, LlamaIndex, OpenAI, Anthropic, AWS Bedrock. Quality, Security & Ops Tested deliverables (unit/integration), structured logging and tracing, env isolation, secret management, access controls, rollout/rollback plans, and documentation for maintainers. I prioritize reliability, observability, and cost transparency. How I Work Architecture → milestone plan → rapid MVP → evaluations & KPIs (accuracy, latency, cost) → iterate → docs & recorded handoff. US timezone; proactive, plain-English updates; async-friendly collaboration via GitHub/Jira/Slack. Next Step Whether you need a hands-on AI engineer or a Tech Lead to guide a team, I’ll help you ship production-grade systems with clarity and speed. Share your use case (data sources, target users, success metrics), and I’ll send a scoped plan within one business day.