Yash Choudhary profile photo 1 on Tal

Yash Choudhary

Software Engineer @PhonePe | Backend & Distributed Systems | IIT Dhanbad EE’25 | Knight @LeetCode 2065 | Expert @Codeforces 1634 | 5 ⭐️ Codechef | Kaggle Competition Master | Ex AI Research Intern Lexsi Labs

Bengaluru

4+ years of work experience

PhonePe
Jun 20251 yr 6 mosPresent

Bengaluru

• Architected and deployed Battery Management System, a real-time Kafka event processing pipeline from scratch,consuming 30M+ daily heartbeat events at ∼350 RPS from 3M+ devices, enriching telemetry via PinotDB, and enabling proactive battery degradation detection and health alerts for merchant devices. • Designed and implemented an inventory and device state management workflow using Flipkart’s DataBuilder framework with async RMQ processing and Airflow/Spark reconciliation, deployed across 12+ revenue-critical device lifecycle flows, eliminating anomalies for 250K+ merchants. • Enabled high-availability for a Tier-1 service by implementing active-active multi-datacenter deployment with disaster recovery drills, achieving 99.99% uptime and reducing RTO by 95%
Veyrisse Estates
Nov 20242 yrs 1 moPresent

Ahmedabad

Lexsi Labs
Feb 20256 mosJul 2025

Mumbai

Fine-tuned Transformer-based models (BERT, GPT-2) and GANs on diverse datasets (textual, tabular, image) using PyTorch, improving task-specific accuracy by ~18%; deployed inference pipelines on AWS with sub-200ms latency for production serving. Implemented AI explainability (SHAP, Attention Visualization) and safety alignment techniques, evaluated 5+ DL architectures across business use cases, and collaborated with MLEs to ship models from experimentation to stable production reducing model drift incidents by 35%.
Kaggle
Jan 20249 mosSep 2024

FinTech Club, IIT ISM
Mar 20222 yrs 11 mosJan 2025

Dhanbad, Jharkhand, India

Collaboratively organized and managed events of the Fintech Club in Technical Fest of IIT (ISM) Dhanbad Concetto 2022 including Tradathon (40k+ participants) and Businellytics (10k+ participants).
GitHub
@yash52003

Yash’s AI Stack

antigravity
antigravity
claude code
claude code
codex
codex
cursor
cursor
gemini
gemini
github copilot
github copilot

Education

Indian Institute of Technology (Indian School of Mines), Dhanbad

Bachelor of Technology - BTech, Electrical and Electronics Engineering

Indian Institute of Technology (Indian School of Mines), Dhanbad

20214 yrs2025
Indian Institute of Technology (Indian School of Mines), Dhanbad

Minor, Marketing Management

Indian Institute of Technology (Indian School of Mines), Dhanbad

20232 yrs2025

Join Yash and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Yash Choudhary on Tal

Headline

Software Engineer @PhonePe | Backend & Distributed Systems | IIT Dhanbad EE’25 | Knight @LeetCode 2065 | Expert @Codeforces 1634 | 5 ⭐️ Codechef | Kaggle Competition Master | Ex AI Research Intern Lexsi Labs

About

I love continuously learning and advancing my skills in software development. I enjoy exploring new ideas and working on innovative projects. With proficiency in Python, JavaScript, React, Node.js, Core Java, Java frameworks, C++, and Docker, AWS, I am proficient in web development and DevOps. Passionate about creating scalable, efficient solutions and contributing to impactful tech projects. Always eager to learn more.

Location

Bengaluru

AI stack

antigravity

claude code

codex

cursor

gemini

github copilot

Work experience

Software Engineer at PhonePe, Jun 2025 - Present · 1 yr 6 mos, Bengaluru, • Architected and deployed Battery Management System, a real-time Kafka event processing pipeline from scratch,consuming 30M+ daily heartbeat events at ∼350 RPS from 3M+ devices, enriching telemetry via PinotDB, and enabling proactive battery degradation detection and health alerts for merchant devices. • Designed and implemented an inventory and device state management workflow using Flipkart’s DataBuilder framework with async RMQ processing and Airflow/Spark reconciliation, deployed across 12+ revenue-critical device lifecycle flows, eliminating anomalies for 250K+ merchants. • Enabled high-availability for a Tier-1 service by implementing active-active multi-datacenter deployment with disaster recovery drills, achieving 99.99% uptime and reducing RTO by 95%

Co-Founder at Veyrisse Estates, Nov 2024 - Present · 2 yrs 1 mo, Ahmedabad

AI Research Intern at Lexsi Labs, Feb 2025 - Jul 2025 · 6 mos, Mumbai, Fine-tuned Transformer-based models (BERT, GPT-2) and GANs on diverse datasets (textual, tabular, image) using PyTorch, improving task-specific accuracy by ~18%; deployed inference pipelines on AWS with sub-200ms latency for production serving. Implemented AI explainability (SHAP, Attention Visualization) and safety alignment techniques, evaluated 5+ DL architectures across business use cases, and collaborated with MLEs to ship models from experimentation to stable production reducing model drift incidents by 35%.

Kaggle Competition Expert at Kaggle, Jan 2024 - Sep 2024 · 9 mos

Fintech(IIT ISM) Member at FinTech Club, IIT ISM, Mar 2022 - Jan 2025 · 2 yrs 11 mos, Dhanbad, Jharkhand, India, Collaboratively organized and managed events of the Fintech Club in Technical Fest of IIT (ISM) Dhanbad Concetto 2022 including Tradathon (40k+ participants) and Businellytics (10k+ participants).

Education

Bachelor of Technology - BTech, Electrical and Electronics Engineering, Indian Institute of Technology (Indian School of Mines), Dhanbad

Minor, Marketing Management, Indian Institute of Technology (Indian School of Mines), Dhanbad

Public links

github: https://github.com/yash52003