Jul 2021 – Sep 2024 · 3 yrs 2 mos · Bengaluru, Karnataka, India · On-site

• Machine Learning Engineer with expertise in designing and implementing robust data solutions for enhanced efficiency and scalability.
• Setup Lakehouse Architecture:
• Setup realtime data ingestion from Kafka into Hudi.
• Setup batch data loading from Snowflake into Iceberg.
• Writing data pipelines using pyspark & Dagster.
• Deployed entire infra e2e using Terraform on AWS.
• Data Platform Development:
• Engineered multi-stage ELT pipelines in Python using Airflow, executed on Snowflake SQL.
• Established multi-tenant pipelines to promote code reusability, enabling seamless integration of new custom logic.
• SQL Optimization:
• Devised strategies to monitor SQL costs on Snowflake, implementing three optimal algorithms for highly used SQL patterns.
• Achieved 4-5x performance improvement compared to intuitive SQL idioms, shared findings on TDS Medium blog.
• Feature Engineering Framework:
• Iteratively designed a feature engineering framework allowing delta format data consumption and exclusive feature creation on new data points.
• Enabled feature definition and offline feature store population with historical data, simultaneously creating online tables.
• ML Platform Collaboration:
• Collaborated with Data Science teams to deliver an ML Platform, facilitating seamless model deployment for batch inferences.
• Utilized Airflow for hourly model triggering, enhancing efficiency and coordination.
• Serverless Data Ingestion:
• Orchestrated end-to-end serverless data ingestion using AWS Glue and Terraform, incrementally ingesting data from Redshift and BigQuery into Snowflake.
• Implemented a control layer in AWS Lambda with Eventbridge scheduler and DynamoDB for state management.
• Integrated GCP components for enhanced cross-cloud capabilities, ensuring a seamless data flow across platforms.

Sep 2020 – Jul 2021 · 10 mos · Bengaluru, Karnataka, India · Remote

• Payments & Checkout Optimization:
• Elevated refund success rates from 60% to an impressive 95%, enhancing customer satisfaction.
• Demonstrated proficiency in microservices architecture while implementing and refining refund flows.
• New Payment Systems Integration:
• Collaborated closely within a cross-functional pod alongside Product Managers and Mobile Developers.
• Successfully onboarded Juspay and Cred Pay, ensuring seamless end-to-end integration with the backend.
• Contributed to the integration process by providing technical expertise and fostering effective communication among team members.

Jan 2020 – Present · 6 yrs 4 mos

• ✦ Qwen OpenAI Agents — AI-Powered Stock Analysis System
• Vibe-coded multi-agent system using Python and DeepSeek in production for NSE/BSE stock analysis
• Serverless architecture with AWS Lambda, SQS, and S3 mountpoint integration
• RSS feed aggregation and AI-powered sentiment analysis for 100 Indian stocks
• Portfolio-level recommendations (BUY/HOLD/SELL) via LLM reasoning
• FastAPI backend with DynamoDB result tracking
• CLI and web interface for investment due diligence
• ✦ Simple4D Capture — macOS Menu Bar Productivity App
• Vibe-coded Swift/SwiftUI app that integrates with macOS Reminders using EventKit
• Implements 4D methodology (Do, Defer, Delegate, Drop) for task management
• Global hotkey (Cmd+Shift+Space) for instant capture via Carbon API
• Modern async/await patterns with MVVM architecture
• Smart due dates, priority levels, and analytics dashboard
• Secure Reminders database integration

Jul 2019 – Aug 2020 · 1 yr 1 mo

• Trading Systems Innovation:
• Led the development of a user interface for the trading system, combining a JAVA Swing-based legacy UI with a modern React-based web UI.
• Revamped the trading system UX, significantly enhancing the experience for traders firm-wide.
• Efficiency and User Value:
• Implemented improvements resulting in a 2x reduction in time required for order placement.
• Achieved a remarkable 20x reduction in client latency through strategic enhancements to the web server.
• Cross-Technology Integration:
• Seamlessly integrated legacy and modern technologies to create a cohesive and efficient trading system interface.
• Collaborated with traders to understand their needs and preferences, translating them into tangible improvements in the system.

Mar 2018 – Jul 2019 · 1 yr 4 mos

Jul 2017 – Feb 2018 · 7 mos

• Exchange Integration Expertise:
• Successfully implemented the updated data feed specifications for a major exchange.
• Ensured seamless integration by maintaining backward compatibility with older specifications.
• Demonstrated agility in adapting infrastructure to evolving exchange requirements while preserving system continuity.

Jun 2016 – Jul 2016 · 1 mo · Hyderabad

• Worked on core of in-house tick store database which holds company’s financial big
• data. Added feature to create a view to display vector data as scalar. Wrote APIs for
• same in C++, Python, Perl and JAVA.

May 2015 – Jul 2015 · 2 mos

• Designed a library called CircosJS. A JavaScript implementation of the famous Circos library. The library can take data in a standard format and construct chart elements as defined by Circos http://www.circos.ca

May 2014 – Jul 2014 · 2 mos · Bangalore

• Design a quiz application on a NodeJS server. User responses are recorded and analyzed.
• Users mistakes and proficiency in answering questions used to set questions in
• further quiz. C3js based charts for user progress analysis. Frontend used AngularJS
• and MySQL used as database.