Krutarth Chaudhari

About

Full-stack and AI systems engineer who has shipped both enterprise fintech and real-time voice AI products. At NCR, they repaired a broken transaction-fee workflow end to end using React/TypeScript, Lambda, and DynamoDB; at Lillup, they turned a research voice prototype into production infrastructure, cutting latency from 350ms to under 200ms and raising user ratings from 2.8 to 4.1. They also built an LLM side project that uses Claude and GPT to cross-review each other through structured JSON consensus rounds.

Key Strengths

• End-to-end full-stack ownership across frontend, backend, data model, and API design
• Strong TypeScript and React engineering with reusable component library contributions
• Deep AWS serverless and DynamoDB architecture knowledge
• Real-time AI/voice pipeline optimization
• Reduced voice pipeline latency from 350ms to under 200ms
• Improved user naturalness ratings from 2.8 to 4.1
• Production-grade LLM systems design with structured evaluation loops
• Robust reliability engineering using retries, backoff, concurrency limits, and fallbacks
• Data-driven monitoring and regression detection in production
• Balances automation vs human-in-the-loop based on operational risk
• Thrives in ambiguous early-stage environments and creates process from scratch
• Clear technical communication and high-value documentation
• Backward-compatible API changes that avoided downstream disruption
• Delivered operational impact by unblocking merchant onboarding and clearing manual backlog

Experience

Education

Awards

• Winter of Code by MSTC 4.0 Winner
• JEE Advanced: Top 2% of 1.1M candidates (2019)
• JEE Main: Top 0.7% of 1.15M candidates (2019)

Skills

Languages