Open to opportunities

Madhav Khosla

@madhavkhosla

Message

I'm a backend engineer building high-scale payments, messaging, and workflow infrastructure.

India

Message

I'm building high-performance backend systems at Trackk, including a Rust Order Management System and BSE IPO bidding flows with UPI/DP integration. I also built an asynchronous KYC PDF-generation pipeline using AWS SQS and S3.

At HighLevel, I owned WhatsApp workflow infrastructure on Google Pub/Sub, processing up to 50M events per month and 1.7M messages daily with a 99.9% delivery guarantee. I improved write throughput to 5,000+ TPM, reduced p95 write latency by 60%, and created TypeScript workflow libraries adopted by four-plus teams.

Previously at BharatPe, I built payment services handling 55K+ daily transactions and $2M+ daily volume, optimized a billion-row MySQL table from 15-second to 200ms queries, and scaled scheduled processing with Kafka. I also build systems projects in Rust, including KiloDB and a concurrent TCP chat server.

Experience

Work history, roles, and key accomplishments

Current

Software Engineer

Current

Trackk

Apr 2026 - Present (4 months)

Built a high-performance Order Management System (OMS) in Rust supporting SL/SL-M order types with real-time trigger evaluation and order lifecycle management. Engineered BSE IPO bidding feature with UPI/DP flow integration and AES-256 checksum auth.

Rust Order Management Systems BSE IPO UPI KYC

SDE - II

HighLevel

Aug 2024 - Apr 2026 (1 year 8 months)

Owned end-to-end WhatsApp workflow infrastructure on Google Pub/Sub, processing up to 50M events/month. Architected fault-tolerant messaging workflows processing 1.7M messages daily with 99.9% delivery guarantee.

TypeScript Google Pub Sub Circuit Breakers Dead Letter Queues Exponential backoff Event Driven Architecture Cache Invalidation Optimistic Locking

SDE - I

BharatPe

Jun 2022 - Aug 2024 (2 years 2 months)

Built payment microservice processing 55K+ daily transactions at sub-200ms latency, handling $2M+ daily volume with 99.9% uptime. Optimized 1B+ row MySQL transaction table through range partitioning and index optimization.

Java MySQL Microservices Kafka UPI Webhooks Partitioning Indexing

Backend Engineer

SDPlus.io

Jul 2021 - Jun 2022 (11 months)

Created architecture and implementation of 'Trickle', a reusable internal backend framework adopted across teams. Transitioned core APIs to GraphQL, eliminating overfetching and reducing average frontend load time by 40%.

GraphQL MongoDB PostgreSQL Backend API Design Performance Optimization

Backend Engineer Intern

Baaz.live

Feb 2021 - Jun 2021 (4 months)

Integrated Baaz's e-commerce stack with Wix, enabling real-time product sync and livestream selling for over 100 vendors. Enhanced user engagement by building gamified livestream features using Socket.io, boosting session length by 2x.

Socket.IO Wix Gamification

Machine Learning Intern

Amplus Solar

Jun 2019 - Aug 2019 (2 months)

Trained a U-Net based segmentation model to detect rooftops and land parcels for solar installation planning. Created a custom Indian rooftop satellite image dataset to accelerate training and validation of segmentation models.

PyTorch U Net Segmentation Satellite Imagery Dataset Creation