Jan 2025 – Jun 2025 · 5 mos · Pune, Maharashtra, India · Hybrid

• Got promoted as Team Lead for the Caching team.
• Optimized TripStack platform further by the following efforts:
• Architected ML-powered caching achieving 98% price accuracy at 3ms p99 latency, serving 1B+ daily requests using gradient boosting and deep learning models
• Drove 5% booking conversion uplift through strategic cache optimization and ML-powered invalidation strategies, saving $24k/month in infrastructure costs
• Spearheaded company-wide migration from REST/JSON to gRPC: reduced payload sizes by 60%, simplified service evolution, and accelerated cross-team development velocity by 40%
• Led team of engineers, mentored on distributed systems design and ML infrastructure

Apr 2024 – Jan 2025 · 9 mos · Pune, Maharashtra, India · Hybrid

• Architected ML-driven caching solution using random forests, gradient boosting (XGBoost), and deep learning (PyTorch) predicting airline price changes with 98% accuracy at 3ms p99 latency
• Developed gradient-boosted models predicting airline price volatility, reducing API calls by 85% and cutting infrastructure costs by $24k/month through predictive cache invalidation
• Deployed production inference pipelines processing 1B+ daily requests with automated model retraining and CI/CD integration
• Optimized FareShare export service: reduced runtime from 40min to 7min (82% improvement) and memory from 48GB to 250MB (99.5% reduction)

Dec 2023 – Mar 2024 · 3 mos · Sahibzada Ajit Singh Nagar, Punjab, India · On-site

• Developed a cross-platform (Windows, macOS, Linux) Configuration and Communication tool using Rust, React, and Tauri framework to work with Tinymesh RF devices. The tool allows changing device configuration parameters, communication in a Tinymesh network and allows users to analyze signal strength across all RF channels of the device.
• Designed distributed IoT systems handling thousands of connected devices with real-time monitoring and telemetry
• Spearheaded integration efforts of various smart meters with Tinymesh's AMI (Advanced Metering Infrastructure) solution based on 6LoWPAN Tinysix stack.

Sep 2019 – Dec 2023 · 4 yrs 3 mos · Sahibzada Ajit Singh Nagar, Punjab, India · Hybrid

• Independent consulting on ML/NLP applications, IoT systems, and distributed architecture.
• Key Projects:
• Built NLP-powered WhatsApp chatbot serving 10,000+ users during COVID-19 pandemic with symptom checking, vaccine center geolocation, and health information access
• Designed and deployed industrial IoT automation system for plywood manufacturing using C, Rust, and Python, improving operational efficiency and enabling real-time monitoring
• Consulted on ML/NLP applications and distributed systems architecture for regional clients across healthcare and industrial sectors

Jun 2017 – Aug 2019 · 2 yrs 2 mos · Noida Area, India

• As a software developer working in the FX team, I worked on developing an end-to-end trading solution for Forex Trading that provides a seamless integration of the front, middle and back office components, providing a seamless forex trading solution for US and European banks.
• Post October 2018, as part of the Trading Tools team, I actively maintained the Multi-Market Spread Trading Tool, which allows traders to create strategies that trade with risk management profiles; and the
• Execution Engine tool, that can place and manage different types of orders on different market gateways and gateway providers.

Jan 2017 – Jun 2017 · 5 mos · Gurgaon, India

• I worked as part of the Core Transaction Processing team in BlackRock’s Aladdin Product Group. My team was responsible for developing an iOS App known as ‘Corporate Actions Manager’ that was integrated with BlackRock’s BMS platform and provided updates to the traders regarding the corporate actions of the securities they were trading in.