Ahmed Makram Alsaidy
Applied AI Engineer — LLM Agents, Function Calling & Automation
Profile
Applied AI engineer and systems builder focused on LLM agents, structured tool use, workflow automation, and reliable API-backed AI systems. Builds and validates agentic workflows end-to-end, with emphasis on strict schemas, deterministic checks, testing, error handling, observability and data consistency.
Core Skills
- Programming & Data
- Python · JavaScript/TypeScript · SQL · JSON · REST APIs · Webhooks
- LLM / Agentic AI
- LLM APIs · Function calling · Structured outputs · Tool routing · Agent workflows · Prompt/eval workflows
- Backend & Data
- FastAPI · Flask · Node.js/Express · PostgreSQL · Supabase · SQLAlchemy
- Reliability & Delivery
- Docker · Git/GitHub · GitHub Actions CI · Unit/integration testing · Logging · Rate limiting · Failure handling
Selected Engineering Work
LI.FI Ops AI AgentOpen Source
TypeScript AI agent using OpenAI Responses API function calls with strict JSON schemas, LI.FI token/quote APIs, bounded tool loops, exact BigInt amount handling, mocked integration tests, Docker and green GitHub Actions CI. GitHub →
TypeScript AI agent using OpenAI Responses API function calls with strict JSON schemas, LI.FI token/quote APIs, bounded tool loops, exact BigInt amount handling, mocked integration tests, Docker and green GitHub Actions CI. GitHub →
Recheck BackendIn Development
Python/FastAPI backend for shipment-inspection workflows with Pydantic validation, SQLAlchemy persistence, OAuth2/JWT authentication and REST endpoints. GitHub →
Python/FastAPI backend for shipment-inspection workflows with Pydantic validation, SQLAlchemy persistence, OAuth2/JWT authentication and REST endpoints. GitHub →
Wasla WhatsApp Automation ServerClient System
Node.js/Express automation service for Wasla Express shipping operations with Supabase, WebSockets, rate limiting, session persistence, Docker and deployment configuration. GitHub →
Node.js/Express automation service for Wasla Express shipping operations with Supabase, WebSockets, rate limiting, session persistence, Docker and deployment configuration. GitHub →
fungi.aiResearch Prototype
Python/Flask + TensorFlow/Keras image-classification application for fungal identification, connecting an ML workflow with microbiology domain knowledge. GitHub →
Python/Flask + TensorFlow/Keras image-classification application for fungal identification, connecting an ML workflow with microbiology domain knowledge. GitHub →
Experience
Alsaidy AI Labs — Founder & Applied AI Engineer
2026–Present- Building a multi-provider AI orchestration layer across OpenAI, Anthropic, AWS Bedrock and local models, including routing, fallback, usage tracking and provider-failure handling.
- Develop agent/task-delegation workflows that use structured tool calls, JSON-based interfaces and controlled execution for multi-step engineering and automation tasks.
- Create evaluation and telemetry workflows for response quality, latency, throughput and cost; containerize services and use PostgreSQL/Supabase-backed data layers.
- Own architecture, implementation, validation, debugging and iteration, with a strong focus on reliable automation rather than one-off AI demos.
Ahmed Alsaidy Studio — Digital Product & Automation Lead
2016–Present- Turn real client and operational problems into structured digital workflows, management systems and automation requirements; own delivery and stakeholder feedback end-to-end.
- Bring 10+ years of disciplined client work, quality control and requirement interpretation to technical AI/data tasks where consistency and attention to detail matter.
Education
Damanhour University — Faculty of Science
B.Sc. in Microbiology & Chemistry Overall Grade: Very Good
B.Sc. in Microbiology & Chemistry Overall Grade: Very Good
Publication
Cyanoremediation of Water Pollutants: Mechanisms, Applications, and Future Perspectives — Damanhour Journal of Pure and Applied Sciences, Vol. 2(1), pp. 60–73 (2026). Read →