Building intelligent systems that scale
Backend engineer specializing in backend architecture, ML integration, and creating seamless user experiences. Currently exploring scalable architectures.
Tech Stack
Technologies I work with
Portfolio
Featured Projects
Swipe or use arrows to explore my work
Bhanzu Buddy
AI-Powered Educational Chatbot
Built an intelligent chatbot using LLM and RAG to answer student questions from in-house curriculum. Implemented rate limiting and queuing for GPU traffic management, achieving 100% uptime with 80%+ monthly student engagement.
- 80%+ monthly active users
- 100% uptime with microservices
- RAG/OpenAI Assistant pipeline for curriculum
Juno
Data Ingestion & Monitoring Platform
Engineered a microservices-based data ingestion module that transforms nested JSON into tabular format for SQLite storage. Implemented comprehensive API diagnostics for Kubernetes/Docker deployments via Elastic Stack.
- JSON to tabular transformation
- Elastic Stack integration
- CI/CD pipeline automation
Dynamic UI
ML-Powered Widget Placement
Developed a neural network system for dynamic widget placement on mobile devices with 92% accuracy. Optimized ML models using TFLite, reducing runtime by 300% on mobile devices. Co-authored patent for the system.
- 92% placement accuracy
- 300% performance improvement
- Patent co-author
Trading System
Post-Trading Automation Platform
Automated complex trading workflows involving 50+ APIs with input/output dependency management using async jobs. Designed an event loop scheduler for multiple reports using Redis Queue, achieving 80% code coverage.
- 50+ API orchestration
- Event-driven scheduler
- 80% test coverage
Career
Professional Journey
Founding Engineer
Stealth Startup
Leading full product development lifecycle from ideation to deployment. Building rapid MVPs to explore product-market fit for consumer applications.
Tech Lead
Bhanzu • Bangalore
- Architected AWS Lambda functions processing 1000+ daily transactions with zero downtime
- Built regression-based trainer allocation engine, improving demo conversions from 7% to 12%
- Led team of 4 to develop QC tool with 97% accuracy in performance evaluations
Senior Software Engineer
Arcesium • Bangalore
- Implemented post-trading API features for cash flow and caching reports via AWS
- Achieved 80% code coverage across ETL codebase with comprehensive unit tests
Software Engineer
Samsung R&D Institute • Bangalore
- Implemented Neural Network for dynamic widget placement with 92% accuracy
- Optimized ML models with TFLite, reducing mobile runtime by 300%
Research
Publications & Patents
Leaf Recognition Algorithm
ISTA 2016, Springer
Algorithm for retrieving medicinal information from leaf images
Dynamic UI Composition
CVIP 2020, Springer
ML-based approach for adaptive user interface generation
Patent: Dynamic UI Layout
201941038129 A (Co-author)
System and method for dynamic UI layout generation