• Designed FastAPI microservices following Clean Architecture, integrating Neo4j & Firebase, improving scalability. • Built async services with httpx & asyncio, reducing API latency by 35%. • Deployed CI/CD pipelines with GCP (Cloud Run, Cloud Build, Workflows), cutting deployment time by 50%. • Integrated Large Language Models (LLMs) from OpenAI (ChatGPT) and Google (Gemini) to power product features, including automatic hashtag generation from text. • Used LangChain, VertexAI, and Google GenAI SDKs to orchestrate LLM usage. • Designed and applied a reflection-based evaluator–optimizer flow for AI agents, improving output quality and aligning results with business objectives. • Implemented distributed caching with Redis and improved monitoring with OpenTelemetry.

• Designed microservices in Flask/FastAPI with GraphQL schemas, queries, and mutations. • Implemented a JWT-based authentication bridge integrating Enterprise X and newly acquired • Enterprise Y, merging user identities across systems to enable secure single sign-on and cross-platform access. • Used SQLAlchemy with PostgreSQL and yoyo migrations; maintained 85%+ test coverage using pytest. • Created AWS Lambda functions for scheduled automation tasks. • Led backend work in the SkillXP squad, aligning with product goals and managing task distribution. • Implemented auth token for Acsendo in Crehana • Documented technical flows, schemas, and best practices on Notion.

• Deployed containers (RabbitMQ, Redis, Celery, Flower) on AWS and in-house servers. • Integrated pre-commit hooks and linters (black, mypy, flake8) to improve code quality. • Built data capture tools with Selenium, BeautifulSoup, RabbitMQ, Celery, and Redis.

• Rebuilt data capture microservices with Python, RabbitMQ, Redis, MySQL, Elasticsearch; deployed on Kubernetes. • Developed a microservice boilerplate that reduced development time by 60%. • Proposed and implemented remote task management workflows, increasing productivity by 20%.

• Built Angular5 dashboard using Google Maps API for real-time electric bike tracking. • Migrated backend from Yii2 to Node.js, reducing API response time by 30%. • Created Flask-based auto-crop tool using OpenCV and AWS Rekognition.