Timothy Nwokeji

Full Stack AI Engineer with 3 years of building intelligent systems. Specialized in LLM integration and AI agent optimization, delivering products that reduced operational costs by 30-40% and cut user workflows by 15%. Led cross-functional teams to ship scalable applications serving users with 99.9% uptime and sub-200ms API response times.

FullStack Engineer at Niche (2025-11 – Present)

Built an AI-driven Donor Intelligence Dashboard using Next.js, TypeScript, Supabase, and NestJS. The app ingests donor CRM export data via CSV upload, validates and parses it, and presents executive-ready insights through interactive charts and KPIs. It also includes an internal AI assistant that can answer natural-language questions about the uploaded dataset using structured data context.

• Live app:
• Login: Email: nwokejitimoth00@gmail.com || Password: timothy-113
• Repo:

Full Stack Engineer at 59 Apps (2024-11 – 2025-08)

Built Native Pay, a P2P mobile wallet on a microservices architecture — a Python/Flask service handling transfers and bill payments, and a Node.js service managing user balances and transaction history, both independently deployed and communicating via shared API contracts. Leveraged atomic transactions to guarantee data consistency across financial operations, integrated tier-1 fintech APIs, implemented KYC compliance, and designed the system to handle 300 concurrent users with zero-downtime infrastructure achieving 99.9% uptime.

• Demo:
• Repo:
• Engineered a production-grade authentication system for a multi-tenant dropshipping platform, Dropshipping Wawi, replacing a stateless JWT implementation with a secure, session-based architecture. Delivered cookie-based auth with rotating refresh tokens, CSRF protection, token versioning, and centralised multi-device logout using atomic database transactions. Implemented a full MFA flow covering setup, verification, and login with TOTP. Built a silent token refresh mechanism on the frontend that transparently retries failed requests after token expiry, preserving user sessions without interruption. Covered the system end-to-end with unit, integration, and Playwright E2E tests including full MFA lifecycle scenarios.

FullStack AI Engineer at PolyQuan Technologies (2024-03 – 2024-08)

• Built an AI SVG generator integrating with OpenAI's DALL·E image generation model, cutting overall design time by 30% by automating SVG creation for both prototypes and final assets. Engineered responsive NUXT frontends and Python backends for relentless performance. Demo:.
• Built a customizable LLM Experimentation platform to help data scientists explore how large language models respond to varying inference parameters (e.g., temperature, top‑P, max tokens). Integrated backend LLM inference to enable side‑by‑side comparison of model behaviors, reducing prompt testing cycles from hours to minutes, and accelerating model evaluation workflows by 80%. Live App: | Repo:

High School Diploma in Science-focused curriculum – Flying High School (2014 – 2020)