Tapas Das
Software Engineer crafting high-throughput distributed systems
About
Backend-focused software engineer with 3 years of experience building systems that operate at scale—5M+ requests per day, 20M+ record ETL pipelines, and 40+ microservices. Specializing in Spring Boot, Redis, and Azure cloud infrastructure, with a track record of turning performance bottlenecks into measurable wins: 80% fewer database calls, 50% lower heap usage, and $15K–$20K in annual cost savings from a single optimization. Equally comfortable diagnosing production memory leaks, architecting multi-agent AI systems, and mentoring engineers to double test coverage.
Education
Indian Institute of Technology (BHU), Varanasi
Aug 2019 – May 2023Bachelor of Technology · CGPA 8.05
Experience
Software Engineer
Medline Industries Pvt. Ltd. · Pune- Designed an in-memory Bloom filter that cut redundant negative SKU lookups, slashing Redis and database calls by 80% and saving roughly $15K–$20K per year while sharpening response times.
- Tracked down a critical memory leak in Spring Security's token storage layer, swapping per-user caching for a shared token model—wiping out 200K+ stale entries and halving heap consumption to stabilize production.
- Built a catalog service handling 5M+ daily requests with cache pre-warming and event-driven invalidation, delivering sub-200ms latency with near real-time data freshness.
- Owned an SKU recommendation engine for in-house brand promotion that drove a 30% sales uplift over six months, backed by daily data pipelines and cache-aside reads for speed.
- Created a reusable Spring Boot starter standardizing RestTemplate, WebClient, Redis caching, and JWT auth—eliminating 30% boilerplate across 40+ microservices and shaving two weeks off each sprint.
- Optimized Talend ETL pipelines to process 20M+ records through multi-threaded batching, boosting throughput by 60% while maintaining 99.95% data consistency.
- Developed a multi-agent AI system (Factory.ai) using JSON-based context handoffs, reducing memory usage by 70–80% and token consumption by 40–50%.
- Shipped Lettuce, a Java CLI for 10+ Azure Redis instances featuring pattern-based queries, deletions, and 8+ custom commands—cutting developer and QA effort in half and plugging into AI workflows via MCP.
- Coached 5 junior engineers on testing practices, lifting coverage from 40% to 85%, and led load tests at 10K concurrent users and 50K RPS to validate peak reliability.
Software Engineering Intern
AI Mage Pvt. Ltd. · Bangalore- Stood up a production Shopify backend on AWS with 5 API Gateway endpoints and 3 Lambda functions, wired with automated CI/CD through CodePipeline across multiple environments.
- Worked around API Gateway's 29-second timeout by chaining asynchronous Lambda invocations for reliable long-running execution.
- Engineered a distributed batch coordination system on DynamoDB using an atomic decrementing counter and real-time cancellation flag to orchestrate parallel Lambda workers—replacing the need for AWS Step Functions.
Projects
AI Powered Knowledge Assistant
Aug 2025- Assembled a fully offline, self-hosted RAG pipeline over 5K+ Confluence documents (1–2M tokens) using ChromaDB and quantized Llama 3.1 8B—guaranteeing complete data privacy with zero external API calls.
- Packaged the system as a containerized microservices architecture with 4 isolated services (LLM, embeddings, vector DB, API), hitting sub-200ms retrieval latency with modular scalability.
- Leveraged bge-m3 embeddings to improve answer relevance by 35–50% over keyword search on internal evaluation sets.
Deployment Tracker
Jun 2024- Created a deployment intelligence platform aggregating AKS metadata for 130+ microservices and MFEs, giving teams clear visibility into releases across environments.
- Automated metadata ingestion through Jenkins and Azure Functions, persisting structured records in Cosmos DB.
- Built a React dashboard with timeline and calendar views for real-time deployment tracking and reporting.
Skills
Languages
Java, C++, JavaScript, SQL
Frameworks
Spring Boot, Spring Security, JPA, React, Express, Mockito
Cloud
Azure (AKS, Service Bus, Functions, Cosmos DB, Azure SQL)
Databases
Redis, PostgreSQL, Cosmos DB
DevOps
Docker, Kubernetes, Jenkins, Git, Bitbucket, SonarQube
Tools
Jira, Talend, Autosys