AI Engineer, Backend Developer & Creator of Autonomous Agents 📍 Dakar, Senegal (Zone 01 Talent / Sonatel Academy / UCAD)
My mission: Solve complex problems by combining cutting-edge Artificial Intelligence (LLMs, RAG, Agents), Data, and high-performance software engineering.
🔗(https://github.com/secke) | 🐦(https://twitter.com/Eseck_K) | (https://www.linkedin.com/in/secke-keita/) 💼(#) | ✉️(#)
The market is shifting towards applied generative AI and software performance. Here's how I can help:
- 🧠 Artificial Intelligence & LLMs: Model fine-tuning (Llama 3.2), advanced RAG architecture, LlamaIndex, Document processing (Docling).
- 🤖 AI Agent Engineering: Designing autonomous agents to automate complex tasks (Web Dev, Quantitative Finance).
- ⚙️ Languages & Performance: Python (Data Science, ML, Backend), Rust (High-performance, Edge AI with the Candle framework).
- 👁️ Computer Vision: Real-time image recognition.
(These projects demonstrate my ability to work with today's most in-demand technologies)
Autonomous agents are the future of applied AI.
- (https://github.com/secke/fullstack-web-dev-Agent): An AI agent designed to assist and automate fullstack web development workflows. This project showcases my ability to build assistants that generate structured code and handle the complexity of modern development.
- (https://github.com/secke/quantitative-financial-agent): An agent applied to quantitative market finance. It illustrates my skill in linking analytical AI with complex financial data for decision support.
Turning raw data into business intelligence.
- (https://github.com/secke/RAG_Assistant_with_Docling_LlamaIndex): Building an advanced Retrieval-Augmented Generation system. Uses LlamaIndex for semantic orchestration and Docling for document parsing. A critical skill for companies wanting AI based on their own data.
- (https://github.com/secke/fine-tune-llama3.2): Training and fine-tuning the cutting-edge Llama 3.2 language model for specific tasks. This shows that I don't just call APIs—I adapt open-source models.
AI shouldn't just be smart; it must also be fast and lightweight.
- (https://github.com/secke/rust-candle-chatbot): A chatbot developed in Rust using the Candle framework (created by Hugging Face). This project is a standout on a résumé: it proves mastery of low-level systems languages combined with machine learning concepts, ideal for ultra-fast inference and edge computing.
- (https://github.com/secke/Real-Im-Recognition): Implementing computer vision pipelines to identify and analyze real-time image streams in Python.
- Zone 01 Talent / Sonatel Academy - Elite training in software engineering (Peer-to-Peer pedagogy, architecture, algorithms).
- UCAD (Université Cheikh Anta Diop) - Strong academic foundations.
Today, many developers know how to hit the OpenAI API. My profile stands out through real AI engineering: I can fine-tune open-source models (Llama 3.2), build robust RAG architectures that don't hallucinate, and code inference in ultra‑performant languages like Rust. Whether it's creating autonomous agents or computer vision solutions, I design production-ready systems.
💬 Ready to innovate?