My work experiences across different companies and roles.
• Revamped research agent architecture, reducing API calls by 67% (from 3 to 1 per query) and saving over $20,000 per month in inference costs, while cutting per-request cost by 90% and improving response latency by 75%.
• Built a high-scale async research engine using Python AsyncIO with 150 concurrent workers, SQS batch processing (100/msg), Redis distributed locks, exponential backoff retries, DAG-based orchestration, and multi-level caching; processing millions of datapoints weekly.
• Deployed three specialized research pipelines tailored to distinct query categories, increasing retrieval relevance by 22%, improving factual coverage, and boosting overall answer quality across production workloads.
• Engineered cross-service credit deduction pipeline across three microservices through checkpoint-based aggregation and idempotent upserts, eliminating duplicate charges while synchronizing balances within 60 seconds.
• Devised a comprehensive LLM evaluation framework featuring a two-stage "LLM-as-Judge" architecture with Perplexity Sonar Pro for web fact verification and GPT-5 for structured evaluation, enabling parallel A/B benchmarking across 8 model variants and cutting manual evaluation effort by 80%.
• Delivered 15+ production enhancements to React.js interfaces for job recommendations and worker dashboards, reducing First Contentful Paint from 1.4s to 0.9s and Time to Interactive from 2.4s to 1.6s, reaching a 92/100 Google Lighthouse score.
• Integrated REST APIs for job and worker matching, processing 15K+ requests/day with an average response latency of 120 ms.
• Spearheaded development of Siamese Neural Network in PyTorch for patient verification system, achieving accuracy rate of 98.6% through rigorous testing and validation of core ML components across 5,000+ patient records.
• Engineered backend services with Node.js and Express.js, designing hospital system integration architecture that enabled 400+ daily patient transactions while maintaining 99.5% system uptime.
built by shiv ratan
© 2026. All rights reserved.
.jpg&w=256&q=75)
