$ whoami
Adithya K
BTech Student | AI Full-Stack Developer
I build full-stack web apps powered by AI — from responsive frontends to scalable backends, wiring LLMs, RAG pipelines, and intelligent APIs into real products.
➜ ~/portfolio git:(main)
npm run build-something-useful
01 — About
Background & focus
I'm an India-based undergraduate (BTech, CSE — Data Science) focused on full-stack AI web development. I like building complete products end to end — clean React interfaces on the frontend, robust APIs and databases on the backend, and increasingly, AI capabilities like LLMs, RAG pipelines, and intelligent agents woven into the middle.
I'm especially drawn to projects that combine traditional software engineering discipline with modern AI tooling — things like MCP servers, vector search, and LLM-powered workflows — rather than AI as a bolted-on feature.
Education
BTech, Computer Science & Engineering (Data Science)
CMR University
Expected Graduation: 2027
02 — Skills
Toolbox
Languages & Frontend
Backend & Data
Auth & State
AI / ML Tooling
Tools & Infra
03 — Projects
Selected work
RepoPilot — MCP-Based Issue Triage Platform
AI platform that helps open-source maintainers triage GitHub issues: auto-categorizes issues, detects duplicates, drafts replies, and generates release notes — all with human approval required. Ships a RAG chatbot and 5 custom MCP tools so any AI assistant can use its capabilities directly.
CivicFix — AI-Powered Civic Issue Management
Lets citizens report civic issues (potholes, garbage, broken streetlights, water leaks) with photo evidence and GPS location. Uses AI to auto-classify complaints, detect duplicates via text and image similarity, route issues to the right department, track SLA deadlines with automatic escalation, and gives citizens a public transparency page to track resolution progress.
CareQueue — Privacy-Aware Clinic Queue and Follow-Up System
Full-stack clinic management platform replacing paper and spreadsheet workflows with digital appointment booking, token-based queue management, and follow-up tracking. Serves four roles — patients, doctors, receptionists, admins — each with permission-scoped dashboards. Includes an AI assistant that answers administrative questions via RAG, explicitly restricted from giving medical advice.
Codebase Q&A — AI-Powered RAG Assistant
Full-stack RAG app that lets developers ask natural-language questions about any public GitHub repo and get source-cited answers grounded in the real code — no more grepping through hundreds of files.
04 — Experience
Timeline
Jul 15 — Aug 7, 2025
AI & Cloud Virtual Internship
Edunet Foundation × IBM SkillsBuild (in collaboration with AICTE)
Credential ID: INTERNSHIP_1748937226683eaa0a58abc
- › Completed a 4-week internship in Emerging Technologies (AI & Cloud), working hands-on with the IBM Cloud Platform.
- › Built practical familiarity with cloud-based AI tooling and workflows as part of a structured, mentor-guided curriculum.
2025
Open Source Contributor — Wagtail CMS
Wagtail (Django CMS)
- › Fixed a server-startup crash (Issue #12791) caused by a circular import: a ModelForm used as creation_form_class on a ChooserViewSet, referenced via a StreamField block, triggered a model import before Django's app registry had finished loading.
- › Diagnosed the full dependency cycle (Page model → StreamField block → ChooserViewSet → ModelForm → model) and evaluated multiple fixes, including dotted-path string support for creation_form_class.
- › Implemented a backwards-compatible fix by deferring form class resolution to runtime instead of import time — merged into Wagtail core (PR #13985).
05 — Contact
Let's build something
Open to internships, freelance work, and collaborating on interesting AI/full-stack projects. My inbox is always open.