# Tanzim Hossain — Comprehensive Technical Dossier & Portfolio Context > Backend-Focused Software Engineer | Dhaka, Bangladesh > Canonical URL: https://0xtanzim.dev ## Professional Summary I'm a software engineering student and backend-focused full-stack developer. I build backend services and complete products across web, mobile, and desktop, including browser extensions and developer tools. My engineering interests span system design, performance, concurrency, and the layers beneath application code: programming languages, runtimes, memory, and operating systems. I also engineer AI-native applications and agentic workflows—bridging local/browser models (Chrome Built-in AI), deterministic backend verification, and autonomous agent orchestration (LangGraph, Google ADK). I take a problem-first approach and choose tools according to what the system needs. ## Work Experience ### Backend Engineer & Mentor — Softsasi (Part-time) *2025-06 — Present | Remote* - Designed backend services, API contracts, and PostgreSQL data models - Built modular REST APIs with clear domain boundaries - Contributed to architecture reviews and design decisions - Mentored developers through code reviews and engineering guidance ## Complete Technical Skills Inventory - **Languages:** TypeScript, Python, Java, Go, C++ - **Backend Frameworks & Runtime:** Node.js, NestJS, Spring Boot - **Database & Storage:** PostgreSQL, Redis, MongoDB - **System Design & Reliability:** Concurrency Control and Transaction Safety, Scheduling and Background Processing, Caching and Rate Limiting, Performance and Reliability Trade-offs, Distributed Systems Concepts (learning) - **Architecture & Boundaries:** System Design (HLD & LLD), Modular and Layered Architecture, Service Boundaries and API Contracts, Multi-Tenant Architecture, Authentication, Authorization, and RBAC, Systems Programming with Rust and Assembly (learning), Framework and Runtime Internals (exploration) - **AI & Autonomous Systems:** Agentic Workflows & Multi-Agent Systems (LangGraph, Google ADK), Model Context Protocol (MCP) & Tool Orchestration, Retrieval-Augmented Generation (RAG) & Vector Search, On-Device Browser AI (Chrome Built-in AI, Gemini Nano), Function Calling & Structured Outputs Validation, Deterministic Verification & Evaluation Boundaries, Agentic Coding & Autonomous Engineering Workflows - **Infrastructure & Cloud:** AWS EC2, AWS RDS, AWS S3, Load Balancer, Docker ## Flagship Projects & Architecture ### Project: Salon Management SaaS - **Role / Status:** Active (Featured: Yes) - **Overview:** Multi-tenant salon booking platform, built solo end-to-end. 27 domain modules across 110+ tables. Booking writes are protected by a pessimistic row lock plus a partial unique index as a hard backstop; a separate Redis mutex guards the temporary slot-hold cache. Caught and fixed a real production billing bug where two disconnected settings quietly disagreed on the platform's cut of every booking. Live at donebyme.dk. - **Deep Dive:** I built this solo — Next.js frontend, NestJS backend, deployed on AWS. Every database table carries a tenant_id, and a middleware injects it on every query. Forget to set the tenant context and the framework throws before the query runs. That's the isolation model: enforced structurally, not by convention. The booking engine supports three ways to assign staff. Auto-assign picks the best person using a weighted score: specialization match, years of experience, seniority tier, and same-day workload. Single staff lets the customer choose one person for everything. Per-service mode lets them pick different staff for each service — the timeline computation engine scans every 15-minute interval from open to close, chains services with buffer times, and checks working hours, existing bookings, and in-flight overlap in one pass. No database calls during the scan. Just math. Double-booking protection on the write path is two real layers, not three — see Key Decisions for exactly which mechanism guards what, since it's tempting to overstate what a supporting Redis lock actually does. Staff are modeled like LinkedIn: one global professional profile, then per-salon employment records, then per-service competencies. A stylist who works at two salons has two employment records, each with different roles, pay rates, and visibility settings. The same person can be a manager at one salon and staff at another. Customers can leave reviews with ratings and photos. The loyalty system tracks points per customer wallet, using idempotency keys and Postgres advisory locks on redemption so concurrent point-spending can't double-spend the same balance. Payments run through a single platform Stripe account. A dedicated table logs every webhook event by its Stripe event ID before processing, so retried webhooks are recognized and skipped instead of double-charging or double-crediting. That table caught a real billing bug, detailed in Key Decisions: two disconnected settings disagreed on the platform's cut of every booking, and the calculated fee was silently discarded before it was even saved. The codebase follows a service-repository pattern. Business rules live in services. SQL lives in repositories. - **Architecture Tracks:** client - **Primary Technologies:** TypeScript, Next.js, NestJS, PostgreSQL, Drizzle ORM, Redis, Google Maps API, PostGIS, Stripe, Capacitor, AWS EC2, Nginx, pnpm Turborepo - **Verified Impact:** Modules: 27 domain modules, DB Tables: 110+ tables across 27 modules, Role tiers: 7-tier RBAC (4 platform + 3 tenant) ### Project: FileTree Pro - **Role / Status:** Active (Featured: Yes) - **Overview:** File tree generator and repository context exporter for VS Code and Cursor. 82,000+ total downloads (42k+ on VS Code Marketplace, 40k+ on Open VSX) with a 5.0 rating. Conflict-aware by design — existing files are never silently overwritten. - **Deep Dive:** A focused developer extension for VS Code and Cursor to turn a described layout into a file tree, preview it, export structured context for LLM prompts, and materialize folders and stub files into the workspace. The goal is a boring, predictable utility: strict validation, no silent overwrites, and a workflow that matches how engineers actually structure codebases. The extension works from a strict, parseable tree format. Parsing rejects malformed input instead of guessing, and the file writer is conflict-aware — existing work is never overwritten by default. Everything the extension generates goes to an unsaved editor tab first; nothing touches disk until you explicitly apply it. Under the hood it's a small, real service-oriented codebase, not a single-file script: a lightweight dependency-injection container wires together a tree-builder service, a file-system service, an exclusion service (glob-based ignore patterns), and four output formatters (ASCII, Markdown, JSON, SVG) behind a common interface. An LRU cache backs the file-system scan so repeated previews of the same folder don't re-walk the disk. 243 tests currently pass across the suite. - **Architecture Tracks:** product, open-source - **Primary Technologies:** TypeScript, VS Code API, Cursor Support, Webview UI, Open VSX Registry - **Verified Impact:** downloads: 82,500+, rating: 5.0 / 5.0 ### Project: NextRush - **Role / Status:** In Progress (Featured: Yes) - **Overview:** A TypeScript backend framework monorepo — 20+ packages covering a segment-trie router, a tsyringe-wrapped DI layer, Koa-style middleware, and Node/Bun/Deno/Edge adapters. Benchmarked with autocannon against Fastify, Express, Koa, and Hono: 14% overhead vs raw Node.js, and it wins the middleware-stack scenario outright (32,220 RPS vs Fastify's 28,744). - **Deep Dive:** NextRush started as a from-scratch study of framework internals and grew into a real monorepo: 20+ packages under `packages/` (core, router, di, decorators, runtime, types, errors, plus adapters for node/bun/deno/edge, 10 middleware packages, and 6 plugin packages), a Fumadocs-based documentation site, and a benchmark suite comparing it against Fastify, Express, Koa, and Hono. The router is a segment trie — not a compressed radix tree, the two are often confused. Routes are indexed by path segment for O(d) lookup where d is segment depth; static routes get a separate hash map for O(1) lookup, and route executors are pre-compiled at registration time rather than per request. Dependency injection is an explicit, honest wrapper around tsyringe (Microsoft's DI library) rather than a from-scratch container — it adds circular-dependency detection and clearer error messages on top. Middleware follows the same Koa-style onion model as Express/Koa: `(ctx, next) => Promise<void>`, composed once at server-start time. Benchmarking used autocannon across six frameworks (raw Node, Fastify, NextRush, Hono, Koa, Express) on identical hardware, with results committed to the repo as raw JSON, not just summarized in prose. The one place NextRush actually wins outright is the middleware-stack scenario — beating both Fastify and raw Node.js — which says more about pipeline design than about the router. Two scoping notes worth being upfront about: the "under 3,000 LOC" and "zero dependencies" claims in the framework's own docs apply to the core/router/types/errors package group specifically, not the full 20+ package monorepo — the DI/decorators stack has two real runtime dependencies (tsyringe, reflect-metadata). And an internal architecture audit in the repo was self-conducted by an AI coding agent, not a third-party security review — it's a useful engineering artifact, but not independent validation. - **Architecture Tracks:** research, open-source - **Primary Technologies:** Node.js, TypeScript, tsyringe, Bun, Deno ### Project: Axiom Framework - **Role / Status:** Experimental (Featured: No) - **Overview:** A Javalin-inspired JVM framework: 5 Maven modules, a real trie-based router, reflection-based dependency injection with classpath scanning, and a JavaPoet annotation processor that generates transactional wrapper classes at compile time. Benchmarked against Spring Boot with autocannon-style load testing — faster on routing and hello-world, slower on JSON body parsing and query-param routing. - **Deep Dive:** Axiom is a small JVM framework built to understand what Javalin actually trades off under the hood — not a Javalin replacement, but a controlled experiment across 5 real Maven modules: axiom-core (routing, HTTP context, config, DI), axiom-persistence (HikariCP pooling plus a hand-built transaction API), axiom-persistence-processor (a compile-time annotation processor), axiom-archetype (a Maven project scaffolder), and a thin aggregator module. The routing DSL reads like Javalin — `app.get("/users", ctx -> ...)` — backed by a real trie router with `:param`/wildcard segment parsing, not string-split matching. The one part of Axiom I'd call genuinely solid under scrutiny is axiom-persistence-processor: a real `AbstractProcessor` using JavaPoet that generates `$Tx` subclasses wrapping `@Transactional` methods in `Transaction.builder(dataSource)...execute(...)` calls — verified with the `compile-testing` library, which asserts on the actual generated source text, not just that compilation succeeds. Dependency injection went through a design pivot worth being upfront about — see Key Decisions and Challenges for the full story, including a docs claim that didn't survive the pivot. Benchmarking against Spring Boot with the same load-testing approach used for NextRush showed a mixed, honest picture: Axiom wins on hello-world and path-param routing, Spring Boot wins on JSON body parsing and query-param routing. A `FAIRNESS_VERIFICATION.md` in the repo documents catching and fixing an earlier benchmark bug (an unfair 3-middleware-vs-1 setup) before publishing results — the kind of self-correction that matters more than the headline number. - **Architecture Tracks:** research, open-source - **Primary Technologies:** Java, Maven, HikariCP, JavaPoet ## Hackathon Sprint Field Logs ### LunaMission — NASA International Space Apps Challenge 2026 (Sep – Oct 2026 (2-Week Build & R&D)) - **Outcome:** submitted - **Summary:** Architected LunaMission end-to-end: spent 5–6 days on intensive astrodynamics and dataset R&D followed by nearly two weeks of systems engineering. Built with Team Softsasi, uniting NASA LOLA 20m laser altimetry with JPL Horizons ephemerides to forecast lunar south-pole solar access and Earth line-of-sight. - **Architecture & System Design:** Built with a decoupled spatial-temporal architecture that I designed end-to-end. An offline Python pipeline ingests NASA PDS LOLA DEMs to compute 360° topocentric horizon masks θ_mask(Az). A high-performance NextRush 4.2 backend service on Bun handles ephemeris queries, in-flight deduplication, and Rolle's Theorem event-splitting solvers. A React 19 + Three.js + Cesium client provides a 5-workspace flight deck featuring interactive 360° sky-ribbons, terrain risk matrix, and rover simulation. - **Tech Stack:** Bun v1.2+, NextRush 4.2, Vercel Serverless, Node.js, NASA JPL Horizons Ephemerides, LOLA LDEM_85S_20M, IAU MOON_ME Datum, Brent's Root Solver, React 19, Three.js, Cesium Ion, Tailwind CSS v4, Vite, NASA PDS Archive, PGDA LOLA GeoTIFF, In-Memory Horizon Cache, Bun Test (1,637 tests), Python 3.11 (Science Pipeline), Git, Docker ### ContentChat AI — Google Chrome Built-in AI Challenge 2025 (Oct 2025) - **Outcome:** submitted - **Summary:** Shipped ContentChat AI: a Manifest V3 Chrome extension running on-device Gemini Nano APIs (Summarizer, Prompt, Writer, Rewriter, Proofreader) for zero-latency, private webpage intelligence. - **Architecture & System Design:** Built as a modular Manifest V3 extension. A lightweight background service worker mediates browser events, content scripts extract clean article DOM via Readability algorithms, and an asynchronous chrome-ai adapter interfaces directly with Chrome Nano's window.ai APIs (Summarizer, LanguageModel, Writer, Rewriter, Proofreader). State is managed via Zustand and persisted locally in IndexedDB. - **Tech Stack:** Google Chrome (MV3), Node.js, Chrome Built-in AI, Gemini Nano (Summarizer, Prompt, Writer, Rewriter, Proofreader), React 19, TypeScript, Tailwind CSS v4, Zustand, Vite, IndexedDB, Git, Chrome DevTools ### Climaxplore — NASA International Space Apps Challenge 2025 (Oct 4–5, 2025) - **Outcome:** 1st Runner Up — Mymensingh - **Summary:** Built Climaxplore during the 48-hour NASA sprint: an AI weather risk assessment system merging OpenWeather data with Google Gemini for activity-specific hazard analysis and Mission Control UI. - **Architecture & System Design:** Climaxplore pairs a real-time data ingestion pipeline with Google Gemini 1.5. Weather parameters from OpenWeather and NASA datasets are normalized into a structured schema, fed through an activity-specific risk evaluation engine, and rendered through interactive 2D/3D map layers and an ARIA conversational agent. - **Tech Stack:** Next.js, Node.js, Vercel, Google Gemini 1.5 Flash, ARIA Chat Assistant, React, Tailwind CSS, Three.js / Mapbox 3D, OpenWeatherMap API, NASA Earth Data, Git, GitHub Actions ## Verified Achievements & Recognition ### NASA International Space Apps 2025 — Climaxplore (Team Softsasi) (2025) - **Accolade:** 1st Runner Up — Mymensingh - **Summary:** 1st runner up (Mymensingh site) at the 2025 NASA International Space Apps Challenge with Team Softsasi on Climaxplore — an AI weather risk platform. My focus: architecture support, technical review, and keeping the system design coherent through the sprint. ## Certified Credentials - **Fundamentals of Operating Systems** by Udemy (Issued: 2026-09-20) [ID: UC-b873c83c-e0d1-4a2d-99ce-3a5d1eafb3df] - **PostgreSQL Bootcamp: Go From Beginner to Advanced, 60+ Hours** by Udemy (Issued: May 2026) [ID: UC-5a4e8b55-3031-4ecd-8f11-450f51a96721] - **LangChain: Agentic AI Engineering with LangChain & LangGraph** by Udemy (Issued: May 2026) [ID: UC-7cdc62d6-34f9-4d89-ab5d-696ce6678478] - **Full Stack Generative and Agentic AI with Python** by Udemy (Issued: Dec 2025) [ID: UC-b02c75d4-8494-4001-8ec7-c178f92f7fe5] - **Reactive Accelerator — Batch 1** by Learn With Sumit (LWS) (Issued: Jun 2024) [ID: LWSCTXN-BLDDDSY5] - **Back-End Development and APIs** by freeCodeCamp (Issued: Jul 2023) [ID: tanzimhossain2/back-end-development-and-apis] - **Hands-On AI: Building Agents with the Google Agent Development Toolkit (ADK)** by LinkedIn Learning (Issued: Nov 2025) [ID: 4d0bcd24c175fdfa2d3c69159df7355e14e6cacd9c74cdf5e551f6d4a1adeaa3] - **Getting Started In Spring Development** by LinkedIn Learning (Issued: Oct 2025) [ID: 1f1c4510fcafe13b1f1aea9a18e34c943f097705c40ea892be3b58fdb49f72ed] - **Complete Guide to Java Design Patterns: Creational, Behavioral, and Structural** by LinkedIn Learning (Issued: Oct 2025) [ID: 8d62247d35b64d3888e6e765cdd59c67fbd00214699e8fd77c298c47cb947e61] - **Introduction to Back-End Development** by Meta (Issued: Aug 2023) [ID: CFWJ2FQJZFYV] - **Introduction to Front-End Development** by Meta (Issued: Aug 2023) [ID: EXT7ERTXGK86] ## Direct Contact & Availability - Primary Email: tanzimhossain2@gmail.com - WhatsApp / Phone: 01838647383 - GitHub: https://github.com/0xtanzim - LinkedIn: https://linkedin.com/in/0xtanzim - X (Twitter): https://twitter.com/0xtanzim