# Janhavi Tatkare > Full-Stack AI Engineer | React, JavaScript, TypeScript | Spring Boot & Node.js | AWS | RAG & LLM-Powered Systems| Ex- TCS| MS CS @NJIT Location: New York City Metropolitan Area, United States Profile: https://flows.cv/janhavitatkare Hi, I’m Janhavi, a Software Engineer with around 5 years of industry experience building AI-driven applications and full-stack solutions. I love applying AI to real-world engineering problems and creating products that feel fast, intuitive, and intelligent. Over the years, I have worked across React, Python, Java, and Machine Learning, delivering features end to end from designing user experiences to building scalable APIs and integrating LLM-based capabilities. My 5 years of industry experience along with my Master’s in Computer Science from NJIT helped me develop a strong engineering foundation, adaptability, and a product-focused mindset. I’m looking for opportunities within teams that are building AI-powered products, intelligent platforms, and modern software experiences. Feel free to reach out. You can contact me at janvtatkare33@gmail.com ## Work Experience ### Software Engineer (AI/ML) @ Community Dreams Foundation Jan 2025 – Present • Own end-to-end delivery of full-stack features across React/Redux frontend and Node.js/Python backend services, shipping 20+ production releases and driving a 16% increase in weekly active usage. • Modernize the UI platform by migrating React Select v1 to v5 and establishing a reusable component library in Storybook, reducing UI regressions by 48% and cutting implementation time for new screens by 30%. • Architect state-driven frontend workflows with editable data grids, dynamic filtering, multi-step form logic, and asynchronous validation, improving form completion by 14% and reducing client-side validation defects by 42%. • Improve frontend performance by reducing Largest Contentful Paint from 3.2s to 2.0s via route-based code splitting, lazy loading, memoization, and query caching; decrease JavaScript bundle size by 31%. • Design and version REST APIs with structured errors, pagination, idempotency, and rate limiting; decompose legacy endpoints into 6 AWS-backed services using API Gateway, Lambda, RDS, S3, and ECS, improving p95 latency from 510ms to 285ms. • Reduce monthly AWS spend by 19% through compute right-sizing, RDS index and query tuning, connection pooling, S3 lifecycle policies, and cache-control/CDN strategies, maintaining stable p95 latency and error rates. • Establish CI-gated automated testing with Jest and Cypress in Docker, increasing coverage to 82% and reducing escaped production defects from 10 to 2 per month. • Build production RAG services in FastAPI using LangChain/LangGraph with vector search integration, supporting 300K indexed documents and maintaining p95 retrieval under 170ms with offline evaluation and quality gates. • Implement serverless LLM pipelines on AWS Lambda with retries, idempotent execution, and structured tracing; reduce hallucinated responses by 35% using retrieval grounding, citation enforcement, and regression test sets; lower cost per request by 38% with caching, token budgets, and model-tier routing. ### AI/ML Engineer @ MetLife Jan 2024 – Jan 2025 • Built and operated full-stack platform capabilities for a healthcare product, delivering 25+ backend features and APIs across the SDLC with product, data, and security partners. • Engineered high-throughput Spring Boot microservices and REST APIs integrating with 130+ EHR systems, improving ingestion reliability to 99.95% and reducing data-validation failures by 40% through strict contracts and standardized error semantics. • Designed an API orchestration layer coordinating multi-service workflows and ML/LLM interactions with audit logging, replayability, and idempotency, reducing end-to-end workflow failures by 37%. • Designed and implemented OAuth2/OIDC and JWT-based access controls with RBAC and mTLS-secured service communication; drove threat modeling and vulnerability remediation efforts, cutting critical security findings by 60%. • Improved performance and scalability on Kubernetes with Spring Cloud patterns, Redis caching, and SQL tuning; reduced average API latency from 250ms to 170ms and increased peak concurrency by 45%. • Built an internal GenAI assistant for analytics and incident triage with prompt versioning, guardrails, PII redaction, and evaluation harnesses, cutting time-to-insight by 33%. • Implemented a Dockerized automated testing strategy integrated into Jenkins CI/CD covering smoke, integration, and regression suites; reduced production rollbacks by 72% and improved deployment success rate to 98.5%. • Architected an observability and analytics pipeline using Elasticsearch, S3, Spark on EC2, and structured logging; enabled near real-time anomaly detection over 500TB+ datasets and reduced MTTR from 2.5 hours to 50 minutes. • Provided production ownership during releases and incidents; wrote postmortems, added runbooks and alerts, and mentored 3 junior engineers, reducing onboarding time by 28%. ### Graduate Assistant @ New Jersey Institute of Technology Jan 2023 – Jan 2024 | New Jersey, United States ### Full Stack Engineer @ Tata Consultancy Services Jan 2020 – Jan 2023 | Mumbai, Maharashtra, India • Owned end-to-end delivery of B2B e-commerce capabilities in an Agile environment, shipping 40+ features across Java/Spring Boot services, Node.js APIs, and React/TypeScript UIs. • Built responsive, accessible React/TypeScript frontends and modernized legacy Angular/jQuery surfaces, improving checkout completion by 11% and reducing UI defect rate by 35%. • Implemented real-time operational dashboards using Server-Sent Events (SSE) and async event streams, improving incident visibility and reducing time-to-detection by 25%. • Created a Backend-for-Frontend (BFF) layer in Node.js to aggregate data from downstream Java microservices, reducing client round-trips by 45% and improving page render time by 30%. • Optimized UI performance with code splitting, lazy loading, and advanced state management patterns, reducing Largest Contentful Paint from 3.4s to 2.2s. • Improved MongoDB performance through schema design, indexing, and aggregation tuning, reducing critical query latency by 32% and lowering compute utilization by 18%. • Automated build/test/deploy pipelines with Jenkins and GitLab; containerized services with Docker and deployed to AWS, improving release frequency and reducing rollback incidents by 65%. • Resolved 80+ production issues using CloudWatch, X-Ray, and Datadog, maintaining 99.9% uptime and reducing average bug-fix turnaround time by 20%. ### Software Developer Intern @ ATS Learning Solutions Jan 2018 – Jan 2018 | Mumbai, Maharashtra, India • Delivered 4 internal tools and student-facing features using Java, Spring MVC, MySQL, and React, improving staff throughput by 22% and supporting 1,500+ monthly student interactions. • Assisted senior engineers with debugging, unit testing, and refactoring legacy code to align with updated coding standards and design patterns, helping reduce regression bugs in key modules. • Created simple dashboards with HTML5, CSS3, and JavaScript to surface metrics from a MySQL backend, reducing manual reporting effort for the team by several hours per month. • Improved page responsiveness by optimizing SQL queries, adding server-side pagination, and reducing over-fetching, cutting median load time from 2.4s to 1.5s. ## Education ### Master's degree in Computer Science New Jersey Institute of Technology ### Bachelor of Engineering - BE in Information Technology University of Mumbai ## Contact & Social - LinkedIn: https://linkedin.com/in/janhavitatkare --- Source: https://flows.cv/janhavitatkare JSON Resume: https://flows.cv/janhavitatkare/resume.json Last updated: 2026-04-13