Harshit Joshi
Graduate Research Assistant
Rayne Township, Pennsylvania, United States
Summary
🤩
Rockstar🎓
Top SchoolHarshit Joshi is a research-driven software engineer with 8 years of experience building practical AI and data-driven systems, currently a Graduate Research Assistant at Stanford SAIL and a PhD candidate in Computer Science. He has blended research and production impact at Microsoft (PROSE), shipping neurosymbolic and LLM-based program repair frameworks that advanced formula authoring across Excel and Power Apps. His background spans robust back-end engineering—contributing notable improvements to the popular invoice2data open-source project—and applied ML work from invoice layout analysis to medical entity linking and image segmentation. Comfortable moving projects from prototype to production, he’s repeatedly cut operational friction (e.g., sales automation and latency reductions) while publishing and deploying state-of-the-art research. Based in Pennsylvania, Harshit pairs deep technical rigor with a knack for pragmatic tooling that accelerates developer workflows.
8 years of coding experience
2 years of employment as a software developer
University of DelhiDoctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at Stanford University
English, Hindi
Stackoverflow
36reputation
1kreached
1answer
0questions
Github Skills (11)
data-mining10
command-line-tool10
python10
pytest9
unit-test9
unit-testing9
regular-expression8
pdf-processing7
angle6
trigonometry6
math6
Programming languages (10)
Github contributions (5)
invoice-x/invoice2data
Feb 2018 - Mar 2021
Extract structured data from PDF invoices
Role in this project:
Back-end Developer
Back-end DeveloperContributions:44 commits, 30 PRs, 16 pushes in 3 years 1 month
Contributions summary:Harshit made significant contributions to the `invoice2data` project, primarily focusing on extending its functionality and improving code quality. They implemented new output formats (XML, JSON), added currency data handling, and incorporated improvements related to date formats, and character encoding. The user also addressed bug fixes, added command-line tool tests, and made code style and dependency updates, demonstrating a solid understanding of the project's core logic and architecture.
pythonpdfstructuredinvoicesdata-mining
duskybomb/invoice2data
Mar 2018 - Mar 2021
Extract structured data from PDF invoices
Contributions:4 PRs, 16 pushes, 12 branches in 3 years
pdf-manipulationpdfstructuredinvoicesstructured-data
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.