Prabhath Palakurthi

Software engineer working on AI, web systems, and developer tools.

I am a software engineer who enjoys turning rough ideas into useful products. I work across AI, web systems, and developer tools, and care most about shipping things that actually hold up in use.

Most of my strongest work has come out of hackathons, internships, and open source, where the focus is on shipping clear, reliable experiences instead of quick demos.

Hackathon projects

TennisIQ, Hacklytics 2026. A system that turns raw tennis match footage into a scouting-style report, with court detection, player and ball tracking, key-moment tagging, and coach annotations. Devpost

Replic, HackPrinceton 2025, top 10 finalist. A tool for early-stage startups to monitor their social channels, draft replies in their own voice, and run them through a lightweight approval flow before anything goes out. GitHub · Devpost

Forensic Fusion, HoyaHacks 2025, top 10 finalist. An AI-assisted memory forensics workflow that helps investigators surface insights from memory dumps and related evidence with a lot less manual sifting. GitHub · Devpost

Watt-Share, HackHarvard 2025. A microgrid simulator for visualizing how surplus solar energy can be shared across a neighborhood in a fairer way. Devpost

EcoHive AI, HoyaHacks 2024. A sustainability-focused platform for eco-friendly ideas, events, and AI-assisted community engagement. GitHub · Devpost

Research

George Mason University, research assistant, LLM watermarking. Studied token-level watermarking at inference time, ran vLLM experiments on decoding and sampling, measured GPU throughput and robustness, and helped maintain benchmarks and documentation.

Open source

chunklint, recently published on PyPI. A static analyzer for RAG chunks that flags structural problems in chunk files before they reach a vector database. PyPI

I also contribute to open source across developer tooling and AI infrastructure. Recent merged work includes adding max_file_size support to the Amazon S3, Azure Blob, and LocalFile sources in CocoIndex.

Experience

TechSur Solutions LLC, full-stack and AI intern, summer 2025. Helped build an internal platform that turns product requirements into Jira tickets tied to Git workflows, and worked on AI-assisted VS Code flows for code suggestions, test generation, and legacy modernization.

Automate365 (VoiceBotics), AI systems intern, spring 2025. Built Python services connected to CRM tools, wired up API and SQL automations, and supported dashboards for monitoring internal workflows.

Rise Consultancy, software intern, summer 2024. Built AWS-backed web apps with React and Node, added serverless APIs, and supported a signal-anomaly detection project on top of cloud data pipelines and ML services.