Headline

Software Engineer @PayPal (Agentic / GenAI focused) | GenAI, ML, Python, Java | M.Tech @ BITS Pilani | Research focused, Results oriented

About

Building autonomous agents—and the required infrastructure required to keep them in check. ​The Engineering: AI Engineer @ PayPal. I sit at the intersection of GenAI and reliable backend infrastructure. My day-to-day involves building scalable, highly accurate web-crawling agents. Because enterprise traffic is unforgiving, I focus heavily on highly redundant architecture design using Python and Java. If a system can break, I assume it will, and architect accordingly. ​The Foundation & Side Quests: B.Tech in CSE from Amrita. Currently pursuing an M.Tech in AI & ML from BITS Pilani while working on a few stealth projects off the clock (meaning I write code for fun after I finish writing code for work). Previous publications in Aerospace CV and Legal LLMs. ​Off the Grid: ​Markets: Overanalyzing the silver super-cycle and running options strategies. ​Machines: Tracking the specs and power-to-weight ratios of naked sportbikes. ​Fuel: Black coffee, Domino's, and unfiltered feedback. ​Keep the corporate fluff; bring me the complex architecture problems.

Location

Bengaluru

AI stack

langchain

langgraph

Browser Use

LlamaIndex

Pinecone

Qdrant

Work experience

SDE - ML at PayPal, Nov 2024 - Present · 1 yr 10 mos, Chennai, Tamil Nadu, India, ◦ Designed and implemented an in-house Agentic web crawling system to analyze PayPal payment option placement across merchant checkout flows and across Chrome, Safari, Android, IOS in an unified intelligence platform, replacing manual competitive intelligence gathering and saving upto 200 Mn $ annually while compressing weeks of works into hours. Analysis identified presentment issues across 11000+ top merchants, led to UX improvements projected to increase conversion by 20% based on A/B test results shared with Merchant Services team. Directly showing an estimated uplift of 9.9 Billion $ USD per quarter in TPV growth as part of Merchant Intelligence team. ◦ Built in-production GenAI automation platform at PayPal that reduced code generation time from weeks to minutes, deployed across 7 countries. Presented system performance metrics to CTO and SVP leadership—developers completing tasks 70% faster with 90% less manual work for mobile app onboarding. Implemented Agentic testing framework using modified version of browser-use and Agent-E that automated 95% of web UI validation workflows previously done manually. ◦ Deployed GenAI tools across Product and Legal teams: Automated market researcher crawling competitor data, PRD generator extracting requirements from documents, and document assembly assistant. Worked with 10+ business units to measure adoption rates and productivity impact—Legal team reduced research time by 60%, Product cut specification writing time by 40%. ◦ Built Model Context Protocol (MCP) servers connecting Claude to internal PayPal APIs—wallet management, transaction history, and developer documentation—reducing context switching time for engineers by 35%. ◦ Recognized for my efforts in H1 2025 Engineer showcase and Q3 2025 Engineer appreciation week.

SDE Analyst at Northern Trust, Aug 2024 - Oct 2024 · 3 mos, Bangalore Urban, Karnataka, India, Worked for a short time on TAM team, focused on optimizing legacy cold storage retreival on snowflake database. Moved on for better opportunities.

Data Science Intern at GE Aerospace, Jan 2024 - Jun 2024 · 6 mos, Bengaluru, Karnataka, India, As a member of the LEAP-Image Analytics team, I contributed to achieving exceptional accuracy in segmentation-based models for the IA workflow, surpassing 92% in F1-Scores and Precision metrics. I successfully developed a state-of-the-art classifier model that achieved an impressive 94.4% F1 Score on the validation set, leading to its integration into the ABM workflow. Furthermore, I helped devise an innovative method for precise quantification of crack detection in aircraft blades, resulting in reduced Turn Around Time (TAT) and lower False Positive/False Negative rates in stress detection models.

Software Development Intern at Northern Trust Corporation, Jun 2023 - Sep 2023 · 4 mos, Bengaluru, Karnataka, India, Developed a dashboard component that significantly reduced the turnaround time for retrieving legacy transactions from over five days to less than fifteen seconds, thereby enhancing the productivity of the Technical Account Manager (TAM). Implemented a robust search logic based on specific requirements and focused on minimizing search latency, improving reliability, and ensuring the security of the dashboard. Successfully achieved comprehensive test coverage for production deployment on company servers in the fourth quarter of 2023.

Machine Learning Project Intern at Wipro, Jul 2022 - Aug 2022 · 2 mos, Bengaluru, Karnataka, India, Created standardized framework for Machine Learning and optimized workflow for Business solutioning in MLops by creating scalable and modular code which helped in training, testing and visualizing machine learning models as well as helping new joiners get comprehensive documentation on entire ML workflow. Reduced KT time from 2 weeks to 4 days.

Education

Master of Technology, Artificial Intelligence and Machine Learning, Birla Institute of Technology and Science, Pilani

Bachelor of Technology - BTech, Computer Science, Amrita Vishwa Vidyapeetham

Public links

github: https://github.com/Hemanthbysani