Computer Science engineering student and full-stack builder focused on AI tooling, automation, applied ML, and developer productivity.
I am building public AI and developer-workflow projects that are useful on their own, not just portfolio filler.
- evidence-first career tooling
- GitHub shipping and repo-quality automation
- practical AI agent and MCP workflows
- full-stack products with clear docs, tests, and deployment paths
An evidence-first career workspace for resume tailoring, saved job targets, application tracking, outreach, interview practice, and analytics.
- Frontend: ResumeAutomator-Frontend
- Backend: ResumeAutomator-Backend
- Live app: resume-automator-frontend.vercel.app
The product is designed to avoid fabricated claims and black-box resume scoring. AI suggestions stay reviewable and are grounded in job-description and resume evidence.
An evidence-first local RAG copilot for source code, Git history, and pull requests. Retrieval-only output is the safe default; generated answers are opt-in and keep immutable source links visible.
- hybrid lexical and semantic retrieval across code and repository history
- a checked browser demo with real source-linked output
- 13 regression tests plus hosted compile and test validation
An authenticated, fail-closed bridge for a laptop-hosted AI inspection prototype, with a stable public gateway, a public application-source snapshot, and a verified v1.0.0 release.
- Ed25519 bridge authentication; no operational records, photos, passwords, or model weights stored on the public gateway
- Node 20/22 CI covering offline health and fail-closed
503behavior - explicit research, dataset, and local-only model limitations instead of production claims
A Node.js and MySQL API that turns public GitHub data into portfolio signals. It includes deterministic regression tests, Node 18/20/22 CI, and a synthetic example response for evaluating the API before setup.
- ship one meaningful improvement at a time
- improve README quality, metadata, examples, and validation
- make responsible open-source contributions where a clear maintainable fix exists
- keep the public portfolio focused on projects with real user value
- Abstractify #77 — built a merged five-agent research orchestrator with staged parallel execution, JSON/SSE responses, memory retrieval, tests, and documentation
- Palinode #89 — aligned Click aliases with command discovery and help behavior
- PARTHA #286 — improved accessible names for repository row actions
Each contribution above was reviewed and merged upstream.
- AI career tooling
- document intelligence and RAG workflows
- browser and developer automation
- repo quality and GitHub workflow tooling
Languages: Python, TypeScript, JavaScript, Java, SQL, C, Kotlin Backend: FastAPI, Flask, Django, Node.js, Express Frontend: React, TypeScript, Vite Data and ML: scikit-learn, TensorFlow, OCR, document AI, RAG workflows Cloud and Tools: AWS, Docker, Firebase, PostgreSQL, MongoDB, GitHub Actions
Back End Developer Intern, WorldInteria.com
Built backend APIs and MVP e-commerce flows using Node.js, React, MySQL, and hosted deployment workflows.
Research Intern, IIEST Shibpur
Worked on blockchain-based certificate verification using Hyperledger Fabric, Go, Node.js, MongoDB, and REST APIs.
- KIITEE Merit Scholarship
- Flipkart Grid 6.0 Semifinalist
- Smart India Hackathon selection
- Machine learning, backend, cloud, and analytics certifications
- Email: dattaaditya74@gmail.com
- LinkedIn: Aditya Datta
- LeetCode: adiya3136
- Codeforces: aditya3136