Saikrishna Vallala

About

Fintech-focused engineer who built an end-to-end KYC verification pipeline for advisor onboarding using Flask microservices, Celery/Redis, and AWS (Lambda/ECS/EC2) with CloudWatch-driven scaling and latency optimizations. Also shipped a production internal knowledge assistant using RAG + embeddings/vector search with guardrails (similarity-based fallback, prompt-injection protections) and an evaluation loop with compliance specialist review that drove measurable retrieval improvements.

Experience

Education

Key Strengths

• End-to-end ownership of regulated fintech KYC/identity verification workflows
• Designs scalable async backends (Flask microservices + Celery/Redis) with production observability (CloudWatch/logs)
• Performance/scalability tuning under load (queue backlog mitigation via horizontal scaling and task decomposition)
• Builds user-facing improvements from feedback (real-time verification status in React dashboard, improved failure handling)
• Ships production RAG/embeddings systems with guardrails (similarity-threshold fallback, input validation, prompt-injection defenses, rate limits)
• Sets up LLM evaluation loops with test sets + human compliance review and measurable metrics (retrieval precision, relevance, citation accuracy)
• Improves retrieval quality through chunking/overlap reindexing based on evaluation results