Question 1

Who is Dibbayajyoti Roy?

Accepted Answer

Dibbayajyoti Roy is a Full Stack Software Engineer based in Agartala, India, specializing in TypeScript, Rust, React, and Next.js. He works at Yupcha Softwares Pvt. Ltd building production-grade SaaS platforms and is the creator of Klinder-OSS, an open-source product analytics SDK.

Question 2

What programming languages does Dibbayajyoti Roy use?

Accepted Answer

TypeScript, JavaScript (ES2023), Rust, SQL, and Python. His primary stack is React and Next.js on the frontend, Node.js and Hono on the backend, and Rust for high-performance systems. He also works with AWS Bedrock, Cloudflare Workers, and Neon Postgres.

Question 3

What hackathons has Dibbayajyoti Roy won?

Accepted Answer

Winner of NITA Arjuna 2.0 National Hackathon (2025, among 200+ teams), Winner of Technovate Project Exhibition (2025), 1st Runner-Up at NITA–ISRO Space Hackathon (2024), and top 500 in the AI for Bharat hackathon (2026).

Question 4

What is Klinder-OSS?

Accepted Answer

Klinder-OSS is an open-source SDK built by Dibbayajyoti Roy that unifies three jobs in a single install: typed event tracking with Zod validation, error-based session recording, and automatic email trigger workflows from product behavior. It replaces the typical PostHog + LogRocket + Customer.io stack. Edge ingestion runs on Cloudflare Workers + Queues; storage is Neon Postgres with Row-Level Security for multi-tenant isolation; the API layer is Hono. A Rust port using workers-rs is in progress targeting sub-10ms p95 latency. Live at klinder-oss.vercel.app.

Question 5

What is diffcore?

Accepted Answer

diffcore is a fast WebAssembly JSON diff engine for JavaScript and TypeScript, built by Dibbayajyoti Roy in Rust and published on npm. It returns real JSON Pointer paths (RFC 6901) and decoded values — not opaque hashes — and emits standard RFC 6902 JSON Patch output, interoperable with fast-json-patch, jsondiffpatch, and any IETF-compliant consumer. It ships applyPatch and revertPatch for state sync, undo/redo, and optimistic UI, includes a React useDiff hook and a CLI, plus a streaming engine for multi-GB files. Benchmarks at 3.3 to 4.1 times optimized pure-JS diff (360–490 MB/s sustained). Runs on Node, browsers, Bun, Deno, Cloudflare Workers, Vercel Edge, Electron, and Tauri. Install: npm install diffcore. Live demo: rust-wasm-library.vercel.app.

Question 6

What is AHTML (Agentic HTML)?

Accepted Answer

AHTML (Agentic HTML) is a shipped sixteen-package npm scope built by Dibbayajyoti Roy at @ahtmljs/* (v1.1.0), plus the ahtml Python SDK, for making web content cheaply and safely consumable by AI agents. From one config a site emits MCP, OpenAPI 3.1, JSON-LD, llms.txt, RSL 1.0, Markdown, and a token-optimal signed snapshot with typed actions carrying explicit cost, reversibility, auth, and side-effects. The scope spans framework adapters (@ahtmljs/next, astro, sveltekit, vite, hono), the contract layer (@ahtmljs/schema) and extractor pipeline (@ahtmljs/extract), the agent client SDK with a dry-run sandbox and universal HTML fallback (@ahtmljs/agent), LangChain loaders (@ahtmljs/langchain and Python), agent-traffic analytics (@ahtmljs/insights), a conformance certification corpus (@ahtmljs/conformance), the AHTML Index registry (@ahtmljs/index), a hosted score badge (@ahtmljs/badge), KV/cache backends (@ahtmljs/kv), a WebMCP bridge (@ahtmljs/webmcp), and a CLI (@ahtmljs/cli) whose init/analyze/score/doctor/mcp commands work on any URL, adopter or not. Measured 4.5–7.3× fewer tokens than raw HTML and fact-extraction accuracy up from 91% to 100% in a 146-run multi-model benchmark; 700+ passing tests. MIT. Repo: github.com/DibbayajyotiRoy/AHTML.

Question 7

What is Learning Copilot by Dibbayajyoti Roy?

Accepted Answer

Learning Copilot is an AI-powered learning assistant built by Dibbayajyoti Roy using AWS Bedrock (Nova Pro and Nova Lite), Next.js, and DynamoDB. It generates structured, level-adaptive explanations with auto-generated D2 diagrams and real-time streaming responses. It placed in the top 500 of the AI for Bharat hackathon and is live at ai-for-bharat.vercel.app.

Question 8

Does Dibbayajyoti Roy have AI and machine learning experience?

Accepted Answer

Yes. Dibbayajyoti Roy has hands-on AI engineering experience including: AWS Bedrock multi-model orchestration (Nova Pro and Nova Lite) with fallback and graceful degradation, streaming inference, prompt engineering, cost-aware architecture with per-mode token budgets, and agent design. He built and shipped a production AI learning assistant placed in the top 500 of the AI for Bharat hackathon.

Question 9

What are Dibbayajyoti Roy's notable projects?

Accepted Answer

Klinder-OSS (open-source product analytics SDK on Cloudflare Workers and Neon Postgres), Learning Copilot (AWS Bedrock AI assistant, top 500 AI for Bharat hackathon), BloodLink (blood donor-seeker matching platform with ~100 pilot users), and EMS (real-time Employee Management System in Rust and React).

Question 10

Is Dibbayajyoti Roy eligible to work in Europe?

Accepted Answer

Yes. Dibbayajyoti Roy is eligible for the Germany Opportunity Card, EU Blue Card, and Netherlands Highly Skilled Migrant (HSM) programme.

Dibbayajyoti Roy — Full Stack & Rust Backend Engineer