The latest articles on DEV Community by Jayesh Thar (@jayeshthar). https://dev.to/jayeshthar https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F2612257%2Fdeda7b23-7599-4a36-9554-f0cc771405bc.png https://dev.to/jayeshthar en Jayesh Thar Thu, 17 Sep 2026 19:15:32 +0000 https://dev.to/jayeshthar/researcher-campus-an-ai-pipeline-that-replaces-5-research-tools-with-one-workflow-4c5o https://dev.to/jayeshthar/researcher-campus-an-ai-pipeline-that-replaces-5-research-tools-with-one-workflow-4c5o <p><em>This is a submission for the <a href="https://dev.to/mlh-hackathon">MLH x DEV Writing Challenge</a></em></p> <h2> What I Built </h2> <p><strong>Researcher Campus</strong> — a full-stack, AI-powered academic research pipeline that replaces 5+ disconnected research tools (novelty checkers, citation managers, compliance checklists, drafting tools) with a single 7-stage workflow: novelty check → gap analysis → manuscript drafting → compliance audit.</p> <p>The core problem: researchers waste hours jumping between tools to check if their idea is novel, find literature gaps, draft a proposal, and then manually verify it meets submission standards. Researcher Campus automates that entire chain.</p> <p>Key pieces:</p> <ul> <li>A <strong>real-time novelty-detection engine</strong> that queries 5 academic APIs simultaneously (Crossref, arXiv, OpenAlex, Semantic Scholar, Europe PMC) and compares results using 768-dimension vector embeddings, with a TF-IDF fallback for offline/degraded scenarios</li> <li>A <strong>5-tier Gemini fallback chain</strong> with auto-failover, so proposal generation doesn't break if one model tier is rate-limited or down</li> <li>A <strong>4-guard compliance auditor</strong> checking citations, anonymity, formatting, and tone before submission</li> </ul> <h2> Demo </h2> <p>🔗 Live: <a href="https://researcher-campus.vercel.app/" rel="noopener noreferrer">researcher-campus.vercel.app</a><br> 💻 Code: <a href="https://github.com/jayesh-thar/Researcher-Campus" rel="noopener noreferrer">github.com/jayesh-thar/Researcher-Campus</a></p> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F8l3cfcw2km14s9pnv2nl.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F8l3cfcw2km14s9pnv2nl.png" alt=" " width="800" height="972"></a></p> <h2> Tech Stack </h2> <ul> <li> <strong>Backend:</strong> 31 REST endpoints, JWT auth, AES-256-GCM encryption, ~7,000 req/sec sustained with 0 errors under load</li> <li> <strong>Infra:</strong> Docker on AWS ECS Fargate, provisioned via Terraform, GitHub Actions CI/CD</li> <li> <strong>Data:</strong> MongoDB + an RDS-based audit-logging layer</li> <li> <strong>Quality:</strong> 21 automated tests, 0 TypeScript errors</li> </ul> <h2> Hackathon Experience </h2> <p>I didn't build this at a live hackathon — Researcher Campus started as a personal problem I kept running into: jumping between five different tools just to check if a research idea was novel. After digging into how existing tools approached this, I decided to build the whole pipeline myself, solo, working online at my own pace.</p> <p>The hardest part wasn't the code — it was the <em>design</em>. Figuring out how to structure a 7-stage pipeline (novelty check → gap analysis → drafting → compliance audit) that actually made sense end-to-end took a long time to get right. I went through several rounds of redesigning the flow before it clicked — moving stages around, rethinking what should run in parallel versus sequentially, and figuring out where the compliance checks actually needed to sit in the pipeline.</p> <p>That one design problem taught me more about system architecture than writing the actual features did.</p> mlhacks ai productivity webdev