Arav Bhardwaj
Full-stack Developer & Founder
Building intelligent systems at the intersection of AI and software engineering. CMU student working on making AI-powered learning accessible to everyone.
About
I'm a developer passionate about building intelligent systems that solve real-world problems. I'm currently studying Statistics & Machine Learning at Carnegie Mellon University, focusing on creating scalable AI-powered applications that make technology more accessible and intuitive.
Right now, I'm building Basics, a startup leveraging AI to bring seamless learning experiences across platforms. Previously, I've architected enterprise AI solutions at KariniAI, built intelligent recruiting tools at Sparo, and conducted medical AI research at Dartmouth-Hitchcock Medical Center.
I specialize in full-stack development with Next.js, React, and Python, alongside deep experience in AI/ML frameworks like LangChain, PyTorch, and LLM integrations. Whether it's building conversational AI systems, computer vision applications, or scalable backend infrastructure, I love creating solutions at the intersection of AI and practical software engineering.
Experience
Co-Founder
BasicsBuilding an AI-powered platform to enable seamless learning experiences across multiple AI platforms. Architecting full-stack infrastructure and developing intelligent learning workflows that adapt to individual user needs.
Fullstack Software Intern
Karini AIDeveloped MCP Registry™ enabling centralized governance and security controls for 50+ production MCP servers, serving 10+ enterprise clients. Architected enterprise conversational AI solution with Microsoft Teams integration, delivering real-time streaming responses and intelligent document processing. Enhanced multi-agent communication architecture by integrating A2A protocol and Ragas evaluation framework.
Backend Software Intern
SparoArchitected automated LinkedIn data extraction pipeline using Puppeteer and Node.js to process 1,000+ professional profiles, reducing manual research time by 85%. Interfaced with connection-mapping algorithms to identify warm networking pathways, improving placement efficiency for AI recruiting platform.
Research Lab Intern
Levy Lab (Dartmouth-Hitchcock)Devised novel GAN architecture in PyTorch to generate high-fidelity synthetic tissue images from colorectal transcriptomics data, reducing dependence on invasive biopsy procedures. Optimized computational workflows using Slurm HPC clusters to process terabyte-scale whole slide imaging datasets, achieving 40% faster processing times.
Projects
Sentinel
FastAPI backend enforcing AI governance policies, cost controls, and compliance monitoring for LangChain applications with <10ms latency overhead. Features policy-as-code with YAML-based declarative rules and RBAC-based multi-tenant isolation.
Ecosort
Autonomous waste sorting system achieving 82% classification accuracy using YOLOv5 model trained on 25K+ TrashNet images. Uses multi-threaded architecture optimized for 30 FPS on Raspberry Pi 5 with precision servo-controlled mechanical separation.
Verity
Blockchain-based platform for AI trust and verification. Full-stack application implementing smart contracts and decentralized verification mechanisms to ensure AI model integrity and transparency.
Liemap
NetLogo simulation model that visualizes the spread of lies in different cultural contexts. Analyzes deception patterns based on lie severity, personal benefit, and social proximity parameters.
PyScan Pro
Learning and debugging tool for students featuring integrated code editor, file explorer, and step-by-step tracing capabilities. Provides comprehensive learning experience for Python programming education.