All Projects

Meeting notes → AI summary, action items, decisions. Full auth, session history, templated prompts. Demo-to-product rebuild.

Drop Amazon links → side-by-side product comparison. Price, rating, features extracted automatically.

Bookmarking API with auto-scraping, tag system, and full-text search. Organised link management for devs.

Standalone JWT microservice — register, login, refresh, revoke. Stateless auth with bcrypt, production-ready security.

Token bucket algorithm from scratch. Configurable Express middleware showing how middleware chains actually work.

WebSocket-based chat with rooms, message history, and online presence tracking. pub/sub pattern in practice.

Background jobs, workers, retries, and failure handling. Async processing patterns you need in every real app.

Short link service with click analytics — referrer, location, time-series data. Redirect logic done right.

Proxy and route requests to different services. Middleware chains, rate limiting, auth — all in one reverse proxy.

Receive, verify (HMAC), queue, and retry webhooks reliably. Handles all the edge cases most tutorials skip.

Pure SQL — no ORM. Joins, indexes, transactions, foreign keys. Understanding what abstractions hide.

Define a model, it generates SQL. Built to understand how Sequelize and Prisma actually work under the hood.

Add Redis to Link Vault — cache invalidation, TTL, cache stampede prevention. When and how to cache.

Up / down / status commands for versioned schema migrations. The infra you need before any real project grows.

One command sets up folder structure, ESLint, gitignore, README. File system ops and CLI argument parsing.

Parse server logs, surface top endpoints, error rates, and slow requests. File streaming and data aggregation.

Pipe git diff to an LLM, get a structured commit message instantly. CLI + AI integration done right.

init, add, commit, log, diff — using SHA1 hashing and file trees from scratch. The most educational project on this list.

HTTP client from scratch — GET, POST, headers. How HTTP works at the socket level, TCP fundamentals.

Search documents by meaning using embeddings. pgvector / Pinecone backend with chunking and nearest-neighbour retrieval.

GitHub webhook triggers LLM review on every PR. Webhooks, GitHub API, and prompt engineering in one project.

Dump notes, links, PDFs — ask questions about them. Full RAG pipeline: chunking, embeddings, semantic retrieval.

Point it at your codebase, get documentation out. AST parsing, code analysis, and LLM prompting combined.

Scrape product prices daily, store history, alert on drops. Cron jobs, time-series data, email alerts in one app.

Fetch any user's repos and visualise contribution patterns. REST API consumption and data aggregation.

Pull from multiple RSS feeds, deduplicate articles using similarity detection, schedule with cron.

Categorise transactions, monthly summaries, spending trends. Data modelling and aggregation queries.

Full SHA-256 implementation using only bitwise operations. No libraries. Understand hashing before using it.

Block creation, chain validation, proof-of-work — pure JS. No libraries, no shortcuts. Just the algorithm.

Build and verify merkle proofs. How blockchain transaction verification actually works under the hood.

Threshold secret sharing as a proper npm package — with docs, tests, and versioning. Crypto meets packaging.

Generate keypair, derive a Bitcoin address from the public key. Elliptic curve cryptography hands-on.

Order matching engine for a decentralised exchange — no blockchain needed. Order books and matching algorithms.

From raw TCP sockets, no frameworks. Parse HTTP requests and build responses from scratch.

In-memory key-value store with RESP protocol. Data structures, persistence, and protocol parsing.

Regex matching and file search without standard libraries. String algorithms and regex engine basics.

Parse commands, fork processes, handle pipes. Process management, Unix fundamentals, and IPC.

Round-robin routing, health checks, failover logic. Networking, proxy logic, and reliability patterns.

Resolve domain names without OS help — UDP, DNS protocol, recursive resolution from the root down.

Watch a directory and trigger actions on changes. Rust file I/O, event handling, systems programming feel.

Same project, different language. Rust networking, threads, and a performance mindset you can't get in JS.

Rust + crypto — directly relevant to blockchain infrastructure work and systems-level engineering.