About
I'm Siddharth, and most places online I go by sidp. I'm an AI and software engineer, currently doing the MS in Artificial Intelligence and Innovation at Carnegie Mellon in Pittsburgh.
Most of what I build starts the same way: something is more annoying than it should be, and I want to know if I can fix it properly. Splitting a restaurant bill is the clearest example. Bill2Go reads a receipt with a vision model, opens a shared session, and lets each person pick their own items in real time. It's live on Google Play and at bill2go.app.
I also like understanding things all the way down. At PES University I specialised in machine intelligence and data science, and my final-year project with three teammates became a paper in Springer's International Journal of Data Science and Analytics, which we presented at IEEE DSAA 2024 in San Diego. At IBM India Software Labs I worked as a software developer for two years. The work I'm proudest of there is a team knowledge base that I proposed, built, and then extended into RAG search over 300+ documents.
Right now I'm most curious about computer vision, LLMs, and how transformers actually work, and I learn best by building something real with them. The question I keep coming back to is how AI can take coordination work off groups of people, so they spend their time deciding instead of organising.
IBM India Software Labs
Software Developer · Jan 2024 to Dec 2025 · Bengaluru (intern Jan to Jun 2024)
Internal work, so there's no public code. These cards are the record.
RAG knowledge base
Proposed and built a knowledge base of 300+ documents for a 70-member team, then extended it into a vector database with embeddings for LLM-based RAG search. Finding an answer went from about a day to seconds.
Differential privacy toolkit
A privacy toolkit for healthcare analytics using OpenDP SmartNoise and diffprivlib, with PII detection and synthetic data generation. Built with a teammate during the internship. It earned Significant Contributor at IBM's 2024 Meraki intern projects.
Enterprise Java tools
Refactored internal Java tools, fixed memory leaks and crashes, and made them about 50% faster. Also built a holiday and leave-scheduling feature (Java/JSP, SQL) used by 70+ people.
Reliability tooling
Spring Boot services for server health monitoring and keystore auto-refresh, plus CI/CD automation that saved about 20 minutes per contribution.
Security testing
Found and reported 20 critical and 50+ other OWASP vulnerabilities in IBM software products, all fixed before release.
Recognition
3rd place at the IBM InCITE 2024 hackathon for a Watsonx Orchestrate MVP.
Earlier
- Software intern, CommunityFarm · built and shipped the K-means order-rerouting recommender into the company's dispatch workflow 2022
- Summer intern, PESU Centre for IoT · vision-driven automated farmbed with a React Native control app 2022
Education
- Carnegie Mellon University · MS in Artificial Intelligence and Innovation (MSAII). Coursework: Computer Systems, Machine Learning, Deep Learning, AI Law and Future Markets. expected May 2028
- PES University, Bengaluru · B.Tech in Computer Science and Engineering, Machine Intelligence and Data Science specialisation. GPA 9.18 / 10. Dr. MRD Scholarship (top 20% of CS students). Best Overall Startup at the Centre for Innovation and Entrepreneurship. 2020 to 2024
Publications
Statistically Inspired Discrepancy Detection for Anomalous Spatio-Temporal Graphs · Tiwari, Chatterjee, Padmakumar, Nair, Das.
- Springer International Journal of Data Science and Analytics, vol. 20, pp. 3267-3281, 2025 · doi:10.1007/s41060-024-00656-0
- IEEE DSAA 2024 extended abstract, presented in person in San Diego · doi:10.1109/dsaa61799.2024.10722835
- ORCID 0009-0008-4910-1511 · the project page
Contact
The fastest route is email. My resume is here as a PDF ↗.