Imran Ahmed
Full-stack developer, ML engineer, and founder of SplitAI. I build software that ships to production — from NLP pipelines to responsive web apps.
SplitAI
AI-Powered Expense Splitting Platform
SplitAI uses AI to make splitting expenses effortless. Describe expenses in plain English, snap a photo of a receipt, or manually enter — our NLP pipeline and OCR system handle the rest. A graph-based algorithm then simplifies who-owes-who into the fewest possible transactions.
NLP Expense Parsing
Describe expenses in plain English — Claude API extracts amount, participants, and split type automatically.
Receipt OCR
Snap a receipt photo. Vision-based OCR with confidence scoring and human-in-the-loop verification.
Smart Debt Simplification
Graph algorithm reduces O(n²) pairwise debts to minimal transfers — 60% fewer settlement transactions.
> About_
Imran Ahmed
Software Engineer · ML Engineer · Founder
Founder of SplitAI. MS in AI at DePaul. I build full-stack apps and ML pipelines that ship to production, not just to GitHub.
React/Next.js, Django, Spring Boot, PostgreSQL, Docker, PyTorch. I've automated away 88% of a QA workflow, built NLP systems that parse natural language at 90%+ accuracy, and I write code that other engineers can actually maintain.
> Projects_
SplitAI – AI-Powered Expense Splitting
NLP pipeline using Claude API to parse natural language expenses into structured JSON with 90%+ accuracy. Graph-based debt simplification reduces settlement transactions by 60%+. Deployed with Docker on DigitalOcean with CI/CD via GitHub Actions.
Handwritten Equation Recognition
Custom multi-headed CNN achieving 87.71% accuracy on 5-token arithmetic equations. Generated custom dataset with PIL and MNIST digits.
> Skills_
[Languages]
[Frameworks]
[Databases]
[Tools]
[Machine Learning]
> Experience_
Software Engineering Intern
Kentish Publishing Company
Mar 2026 – Present
- ›Architecting a full-stack rebuild of a narrative analysis platform, migrating legacy React systems into independent infrastructure with custom backend APIs and relational database design.
- ›Engineering software from documented algorithms, including a proprietary 3-process reflection framework, building RESTful endpoints and React interfaces to serve real-time writing analysis to end users.
Founder & Software Engineer
SplitAI — AI-Powered Expense Splitting Platform
Dec 2025 – Present
- ›Engineered a multi-stage NLP pipeline using Claude API that parses natural language expense descriptions into structured JSON, with prompt optimization achieving 90%+ extraction accuracy on fields like amount, participants, and split type.
- ›Built a vision-based receipt OCR system with model selection analysis (cost vs. accuracy tradeoffs), client-side image preprocessing, and confidence-scored output with human-in-the-loop verification.
- ›Architected a Spring Boot backend with RESTful APIs, JWT authentication, PostgreSQL with JPA/Hibernate, and relational data modeling across users, groups, expenses, and payments.
- ›Implemented a graph-based debt simplification algorithm that reduces O(n²) pairwise debts to minimal transfer sets, decreasing average settlement transactions by 60%+.
- ›Deployed with Docker on DigitalOcean with CI/CD via GitHub Actions, Spring profiles (dev/prod), and managed PostgreSQL.
Software Development Intern
Black Fox Security Solutions • Aurora, IL
June 2024 – Aug 2024
- ›Refactored and debugged Python and JavaScript scripts to improve system reliability and created automation scripts that enhanced testing efficiency.
- ›Collaborated with developers via Git and GitHub in an Agile environment with daily standups and sprint planning.
IT Support Specialist
Humble Properties • Chicago, IL
Oct 2021 – Sept 2022
- ›Resolved 20+ hardware and software issues per week while communicating technical solutions to non-technical users, achieving a 95% satisfaction rate.
> Get In Touch_
Looking for SWE and ML engineering opportunities. Building something interesting? Let's talk.