- 10k+ PDFs RAG Pipeline: Built an end-to-end ingestion and vectorization pipeline, tuning retrieval quality (via LangFuse tracing) to achieve a +22pt improvement in search success.
- 60% Faster Deployments: Architected secure, automated delivery with GitHub Actions, Docker, Trivy, and Cosign, slashing deployment time from 18 to 7 minutes.
- Certified Expertise: AWS Certified Solutions Architect (Professional), OCI AI Foundations (Associate), and OCI Foundations (Associate).
Designing practical RAG workflows with AWS Bedrock, Dify, OpenSearch, Python, and Langfuse. Evaluating model responses and optimizing vector search retrieval. | Building repeatable delivery pipelines with GitHub Actions, Docker, GHCR, Trivy, Cosign, SBOMs, canary releases, and rollback-aware release flows. |
Implementing Datadog metrics, logs, APM, monitors, and LLM observability (LangFuse). Experience with GKE production monitoring and infrastructure-as-code alerts. | Operating public services and private lab workloads across AWS, Azure, GCP, Cloudflare, Vercel, and self-hosted infrastructure with Zero Trust boundaries. |
