WARDAH HAYA
Agentic AI Engineer
& Full-Stack Developer
Professional Summary — I design agentic frameworks, high-throughput asynchronous backends, and multi-modal ML systems. Expert at driving architectural efficiency and deploying award-winning distributed applications.
- 70%
- Fewer agent cycles
- 60%
- Workflow gains
- <1s
- RAG latency
02 · Technical Stack
Tools that ship production-grade systems
A curated stack for agentic AI, async backends, and multi-modal pipelines.
- Python (Asyncio)
- TypeScript
- JavaScript (ES6+)
- SQL
- C++
- Tailwind CSS
- HTML5/CSS3
- LangGraph
- LangChain
- LlamaIndex
- Model Context Protocol (MCP)
- PyTorch
- TensorFlow
- Scikit-Learn
- XGBoost
- Multi-Agent Systems
- RAG
- Fine-Tuning (LoRA, QLoRA)
- OpenCV
- Semantic Search
- FastAPI
- Node.js
- React
- Next.js
- Streamlit
- ReportLab
- WebSockets (Socket.io)
- REST APIs
- JWT
- GCP
- Docker
- CI/CD (GitHub Actions)
- Vercel
- LangSmith
- Jest
- Drift
- Automated Eval Suites
- System Design
03 · Professional Experience
Selected roles, measurable impact
Each engagement shipped agentic infrastructure with quantifiable performance lifts.
04 · Featured Projects
Systems shipped end-to-end
From agentic orchestration to multi-modal inference and distributed marketplaces.
AID
Autism Pre-Screening System
Multi-modal diagnostic pipeline matching a fine-tuned ResNet-50 vision network with an XGBoost classifier. Orchestrated async FastAPI inference, Groq Llama-3.1 structured parsing, and automated ReportLab generation.
- ResNet-50
- XGBoost
- Groq Llama-3.1
- ReportLab
Hunarkada
Hyperlocal Marketplace
Distributed marketplace featuring an algorithmic 0–100 user trust score system, Leaflet.js geofenced mapping, and identity-masked chat via Socket.io. Optimized Next.js and Tailwind CSS rendering lifecycles.
- Next.js
- Socket.io
- Leaflet.js
- Trust Score
05 · Other Projects
More work on GitHub
Explore the full archive of builds, experiments, and academic projects.