I’m a Software Engineer delivering AI-driven automation, full-stack platforms, and cross-functional enablement across EMEA. I thrive on rapid prototyping, clean code, and translating stakeholder feedback into production-ready features.
Holding an MSc in Computer Science (Distinction expected) and a BEng in Computer Engineering, I build across MERN, Python, and AWS stacks to solve complex platform and data challenges.
Skills
Experience Level
Language
Work Experience
Education
Qualifications
Industry Experience
Weed Therapy is an AI-assisted wellness companion designed to support mindfulness, emotional regulation, and cannabis habit management. Users interact with an empathetic AI “therapist,” complete daily check-ins, and revisit past sessions to track progress. The system combines React, Node.js, and MongoDB with Google Gemini to deliver a therapeutic, privacy-focused experience.
Project Link: https://www.twine.net/signin
How It’s Built
Tech Stack:
React, Tailwind CSS, Node.js, Express, MongoDB, JWT Auth, Cloudinary, Google Gemini API, Web Speech API
Frontend
React + Tailwind for a calm, mobile-first UI
Session-type workflows (Crisis, Daily Check-In, Reflection, Habit Builder)
Daily check-in popup with backend state syncing
Speech-to-text message input
Optimized chat experience with auto-scroll, caching, and typing indicators
Backend
Node.js/Express REST API
Mongoose models for users, sessions, and daily check-ins
Secure JWT authentication with refresh-token flow
Cloudinary for avatar/media storage
Robust middlewares for auth, rate limiting, and validation
AI Integration
Google Gemini for emotional, context-aware chat responses
Different prompt templates per session type
Safety rules for sensitive conversations
Core Features
AI Therapy Chat: Adaptive responses tailored to stress, cravings, reflection, or journaling.
Daily Check-Ins: Mood, craving, stress, and notes logged once per day.
Session Types: Structured conversations like General Therapy, Grounding, Craving Help, Reflection, and Habit Building.
Anonymous Mode: Sessions stored without identifiers; UI clearly marks anonymous chats.
Session History: Authenticated users can revisit or continue previous sessions.
Mobile App (WIP): React Native version in development.
Key Optimizations
Selective history fetching to reduce payload size
Centralized session-type registry and UX-first ordering
Per-user daily check-in logic using backend + localStorage
Rate limits and input validation around AI endpoints
Improved performance via lazy-loading, efficient state handling, and streamlined scrolling
- Personalized Tour Planning
- LLM-Powered Conversational Guide
- Interactive Maps
- Smart Routing Engine
- State Persistence
- Streamlit, Folium
- Python, NetworkX, GeoPandas
- Google Gemini API
- OpenStreetMap
- Load a city
- Choose constraints
- Route optimization via Weighted A*
- Conversational AI guidance
- Map visualization
- Transport integration
- Better AI prompt engineering
- Mobile UI
- Deployment pipeline
AI Tour Guide is a prototype AI-powered city exploration assistant that generates personalized, time-optimized tours using real geographic data and conversational LLM guidance.
Features
Tech Stack
How It Works
Current Status
Early WIP prototype.
Planned Enhancements
Hire a App Developer
We have the best app developer experts on Twine. Hire a app developer in Newcastle upon Tyne today.