I turn ideas
into shipped
products.
I solve product problems end-to-end — intelligent automation, conversational AI, cross-platform mobile apps, and cloud infrastructure that scales without idle cost.
Selected Work
Projects I've built and shipped — from a live two-app travel platform to AI-native prototypes.
Wegether
Travel Platform · US-based startup
A two-app travel platform connecting tourists with local guides across multiple time zones. I joined in the early stages, designed the serverless architecture, and built across the full stack for ~5 years — from DynamoDB schema design to App Store deployment.
Business outcome: The serverless architecture avoids idle infrastructure cost — you pay for usage, not uptime. Real-time chat on EC2 with presence indicators means tourists and guides communicate without polling delays. The LLM itinerary builder was the platform's first AI feature, integrated directly into the tourist app's booking flow.
Language Master
AI Language Learning App
An AI-native language learning prototype that replaces static lesson trees with real-time conversational AI. Speech input, pronunciation scoring, and a personalized lesson path — the AI adapts to the learner rather than routing them through a fixed curriculum.
Why it matters: Built to explore how agentic AI patterns (tool calling, memory, multi-step reasoning) behave in a mobile context. The codebase demonstrates the integration points between an LLM backend and a React Native client — the same patterns that power production AI features.
Screenshots from the initial design prototype
This Portfolio
Personal Portfolio Website
A static Next.js site designed for fast, zero-cost deployment. Built with TypeScript, Tailwind CSS with custom design tokens, and Framer Motion for scroll-triggered animations — all compiled to static HTML that runs on any host without a server.
What this demonstrates: the same attention to performance, accessibility, and deployment discipline I bring to client projects. Static export means no runtime costs, no cold starts, and a single build that deploys identically to Vercel, GitHub Pages, or Netlify.
Skills & Stack
Every technology listed here was used in a shipped product — from the Wegether platform to agentic AI systems.
Experience
~5 years building and maintaining a live production platform, plus earlier engineering training.
Wegether Travel LLC
Fullstack Mobile & AI Developer
- Designed and scaled a serverless AWS backend (Lambda, DynamoDB, EC2, API Gateway, S3, SNS, Amplify) serving tourists and guides across multiple time zones.
- Built and maintained two React Native apps — Wegether Tourist and Wegether Guide — each deployed to both iOS and Android (4 app store listings).
- Modeled DynamoDB schemas for travel-platform read/write patterns: bookings, user profiles, reviews, and multi-currency payments.
- Architected a real-time chat system on EC2 with presence indicators and push notifications via SNS — no polling, no delay.
- Integrated an LLM-powered AI itinerary builder — the platform's first AI feature — directly into the booking flow.
- Implemented multi-language support, multi-currency pricing, and split-payout payment processing.
- Set up CI/CD for automated testing and deployment to both App Store and Google Play.
INTTIC Oran
Graduation Project — Desktop App
- Built a desktop email management application using JavaFX for the UI and JDBC + SQLite for data persistence.
- Project theme: Computer-Mediated Communication, Models and its uses.
Algerie Telecom
Network Internship
- Hands-on with regional physical network infrastructure.
Ooredoo Development Department
Trainee — Android Dev & IoT
- Android mobile development fundamentals.
- Arduino & IoT basics.
ATM Mobilis
Network Internship
- Hands-on with live cellular network infrastructure.
Education
Languages
About
I'm Abdelhadi Djafer — an Fullstack Developer & Agentic AI Builder based in Chlef, Algeria.
For ~5 years I was the technical core of Wegether Travel LLC, a US-based startup, where I built and maintained a two-app travel platform across iOS and Android. I took over in the early stages, designed the serverless AWS architecture (Lambda, DynamoDB, API Gateway, EC2, S3, SNS), and integrated the platform's first LLM-powered feature — an AI itinerary builder.
Today I build agentic AI systems — multi-agent workflows, tool-calling pipelines, MCP integrations, and autonomous systems that move beyond chatbots into software that reasons, plans, and acts.
I hold 6 Anthropic Academy certificates covering the full agentic AI stack:
- Claude Code in Action
- Claude Code 101
- Introduction to Subagents
- Introduction to agent skills
- Introduction to Model Context Protocol
- Claude with the Anthropic API
Trilingual (Arabic, English, French). I work remotely, communicate across time zones, and take projects from architecture to deployment.
What I'm focused on
Multi-agent workflows, tool calling, MCP integrations, and autonomous AI products. Building beyond chatbots — systems that reason, plan, and act.
From DynamoDB schema design to App Store deployment. I build entire products, spanning infrastructure, backend, and mobile clients.
Integrating LLMs into real user workflows where AI is the core experience rather than a sidebar feature. Shipping, measuring, iterating.
Let's build something.
If you need a system built — whether an agentic AI feature, a mobile app, or a serverless backend — reach out. I take projects from architecture to deployment.
abdelhadi.djafer.02@gmail.comNext.js · GitHub Pages