VASU BANSAL
Software Engineer & AI Researcher building impactful solutions at the intersection of technology and innovation
About
I'm a Computer Science and Business Administration student at UNC Chapel Hill, passionate about building impactful software solutions and advancing AI research. Currently working as a Software Engineer Intern at RevisionDojo (YC F24) and an AI/ML Engineer at One Media Connect, I've contributed to projects reaching 400K+ users and managed $10M+ in ad spend through AI automation. My work spans full-stack development, AI research, and technical leadership, with a focus on creating accessible and innovative solutions.
Full-Stack Development
React, Next.js, Express.js, FastAPI, Node.js, Python
AI/ML Engineering
LLMs, Reinforcement Learning, Pandas, NumPy, Scikit-learn
Database & APIs
PostgreSQL, MySQL, MongoDB, RedisDB, REST APIs
Technical Leadership
Team management, system design, agile methodologies
Actively Building
Products I'm shipping right now — live, iterating, and seeking feedback.
Phad
Social media can surface consumer behavior shifts before mainstream finance or analysts price them in. Phad measures these trends systematically — tracking acceleration, cross-feed recurrence, and public-market connections.
Eclair
Active recall and spaced repetition are effective, but Anki's review loop can feel dull and easy to game. Eclair keeps the scientifically valuable retrieval loop while changing the experience so students stay engaged and test concepts more honestly.
Projects
Projects spanning AI/ML, full-stack development, and research that demonstrate technical excellence and innovation.
ClassWatch
AI-powered classroom monitoring platform — Grand Award Winner at HackDuke 2026
ClassWatch
AI-powered classroom monitoring platform — Grand Award Winner at HackDuke 2026
MAWSS: Nature Cleaning Nature
Bio-engineered moss panel system for passive CO2 capture and air purification — Claude Builder Club Hackathon Winner
MAWSS: Nature Cleaning Nature
Bio-engineered moss panel system for passive CO2 capture and air purification — Claude Builder Club Hackathon Winner
ClaimMax
AI-powered auto insurance claim generator
ClaimMax
AI-powered auto insurance claim generator
ASCENT: Aerospace System for Chemical Emissions & Numerical Tracking
AI-powered rocket launch pollutant dispersion simulator
ASCENT: Aerospace System for Chemical Emissions & Numerical Tracking
AI-powered rocket launch pollutant dispersion simulator
Independent Research on One-Drop Numbers and Arrow Theory
Mathematical research on numerical sequences and Arrow Theory
Independent Research on One-Drop Numbers and Arrow Theory
Mathematical research on numerical sequences and Arrow Theory
PEWP (Predicting Exoplanet Weather Patterns)
ML model for forecasting exoplanet weather from atmospheric data
PEWP (Predicting Exoplanet Weather Patterns)
ML model for forecasting exoplanet weather from atmospheric data
Experience
Professional experience in software engineering, AI research, and technical leadership.
BCI Gen AI Integration Team Lead
- Lead GenAI integration efforts for the Brain-Computer Interfaces Research Group, translating research objectives into scoped technical initiatives and implementation plans
- Coordinate across research, engineering, and project stakeholders to define priorities, align execution, and keep early-stage work moving through ambiguity
- Support applied AI development for BCI-focused workflows under Prof. Raghavendra Pothukuchi, connecting experimental goals with practical product and system decisions
Software Engineering Intern
- Developed cross-platform SAT prep app serving 15,000+ daily active users using React Native and Express.js
- Reduced API response times by 20% through Redis caching and PostgreSQL query optimization, improving experience for 400K+ student community
- Implemented real-time progress tracking and adaptive flash card algorithms, increasing average daily study time by 35%
AI/ML Engineer & Business Analyst
- Developed AI automation tools for real-time bid adjustment and budget optimization across $10M+ in managed advertising spend across 100+ accounts
- Implementing Mixed Market Models forecasting campaign ROI with 85% accuracy, reducing client cost-per-acquisition by 30% over 6 months
- Developing automated reporting pipeline integrating Salesforce, Google Analytics, and Meta Ads API, reducing manual report creation time by 75%
Chief Financial Officer & Technical Team Lead
- Leading cross-functional team of 20+ developers and designers to build Luminary, a campus accessibility navigation app serving 150+ students with mobility limitations
- Architecting pathfinding algorithm integrating OpenStreetMap data with campus building blueprints, reducing navigation time by 15% in beta testing
- Managing $10K+ budget and overseeing operations for a 500+ member organization, coordinating funding and resource allocation
Team Lead
- Leading the OCSL project in collaboration with OCSL and Chapel Hill police to automate student event registration workflows, translating stakeholder needs into clear product and technical requirements
- Designing and building a FastAPI + PostgreSQL backend to support secure submissions, address verification, and reliable workflow orchestration
- Implementing a real-time administrative review system for event eligibility checks, improving operational visibility and reducing manual review friction
- Coordinating project execution across teammates by aligning scope, assigning priorities, and keeping engineering work tied to client and user needs
AI Research Intern
- Developed reinforcement learning framework using cosine similarity metrics and GPT-4-based feedback loops, improving output fidelity by 74% compared to baseline RLHF models
- Co-engineered Chain-of-Thought reasoning pipeline extracting and modeling user personas from Twitter, LinkedIn, and structured interview data with 92% accuracy across 500+ test profiles
- Authored research paper on novel reinforcement learning framework for AI persona modeling, combining LLM-as-a-judge evaluation with quantitative similarity metrics
AI Research Intern
- Contributed to research on AI and Fully Autonomous Unmanned Aerial Vehicles (FAUAVs)
- Explored applications of FAUAVs for real-time traffic incident detection
- Researched smart city operations, emergency response, and urban planning applications
Let's Connect
Interested in collaborating on innovative projects or discussing opportunities? Let's connect.