Towhid Al Faysal
Full-stack developer with almost 2 years of professional experience building end-to-end web applications with Python/Django and React/Type Script on Postgre SQL, owning features from architecture through delivery. AI-augmented builder: uses Claude and Claude Code daily and built an MCP (Model Context Protocol) server connecting AI agents to production APIs.
Elite competitive programming background (ICPC Dhaka regionalist, 2000+ problems solved), and a problem solver who finds answers where docs and forums run out.
Experience
Software Engineer (Full-Time)
Automation Solutionz
Remote
Built a Python/FastMCP MCP server exposing ZeuZ's project-management domain as typed read and write tools for AI agents, with per-user permission enforcement and audit logging. Designed a stateless, gateway-ready MCP architecture extensible across service domains. Built a Daily Standup reporting system in Django and React with AI-generated summaries, worklog and ticket evidence, Git Hub PR reviewer assignment, and Celery scheduling.
Designed a unified linking API with React Query UI that cut link-tab network calls from 4–5 to 1, enabled same-type linking, and added 54+ API tests. Integrated Snowflake and remote device-farm testing into the Python automation node. Resolved automation and notification reliability issues, including Windows nmap/wapiti report generation, scheduler migration and API regressions.
AI Model Trainer (Coding)
Outlier
Remote
Trained and evaluated AI models on coding tasks, reviewing outputs and providing structured, well-documented feedback to improve code generation, reasoning, reliability, and overall quality.
Education
Chittagong University of Engineering and Technology
Bachelor of Science
Skills
TypeScriptJavaScriptReactHTML/CSSPythonC++CREST API designsystem designDjangoFlaskFastMCPCelerySQLPostgreschema migrationsPydanticunit & API testingLLM integrationMCP servers (FastMCP)model training & evaluationTensorFlow/KerasPyTorchGitDockerCI/CDLinuxGitHub APIsClaude & Claude Codeprompt engineering
Key Achievements
42nd place, ICPC Dhaka Regional Preliminary Round (2500+ teams)
182nd place, ICPC Dhaka Regional Main Round
16th place, CU IT Fiesta 2024 Inter-University Programming Contest
1st Runner-up, CUET Intra-University Programming Contest 2023
Problem Setter for multiple programming contests
Projects
Graph Visualization & Analysis Web Application
GitHub
Designed Flask REST APIs for backend graph analysis (shortest-path via NetworkX) behind an interactive React frontend with pan, zoom, draggable nodes, and step-by-step DFS/BFS traversal driven by state-machine logic. Built a hybrid classification system combining rule-based analysis with a Tensor Flow/Keras model (Tree/Cycle/DAG), and converted a trained PyTorch model to Keras for production inference.
Tech Stack: Python, Flask, NetworkX, Tensor Flow/Keras, React, React Flow, Zustand, Ant Design