Abdelkrim Dakouan

Abdelkrim Dakouan is a AI/ML Engineer specializing in building intelligent, scalable systems for data-intensive platforms. With a strong foundation in API design, recommendation algorithms, and AI-enhanced data processing, he excels in developing robust backend architectures and deploying cutting-edge AI solutions in production environments.

At Valhko, Abdelkrim led the development of a multi-agent AI system powering Arya, an advanced orchestration layer combining reasoning, memory, deepsearch, and planning across specialized agents. He designed coordination protocols and backend infrastructure to enable dynamic task execution, significantly improving automation and contextual understanding our agentic workflows.

Previously at Eliotron, a startup he co-founded, he built REST APIs and recommendation engines for financial and e-commerce analytics, deploying optimized microservices with Flask and WebAssembly (WASM) for performance-critical environments. His backend solutions enhanced operational intelligence and improved trend forecasting using fine-tuned language models like BERT.

At MA, Abdelkrim created a document intelligence system using GraphRAG techniques, combining knowledge graphs, embeddings, and LLMs for real-time querying and response generation from complex PDF documents. His pipeline integrated both structured and unstructured data for improved search accuracy and contextual relevance.

Earlier, at OCP, he developed predictive maintenance tools using machine learning (XGBoost, VAR) and natural language processing, contributing to the company’s IoT-based ISEE Platform for smart equipment monitoring.

He holds a Master’s degree in Data Science and has earned certifications in LLM application development and prompt engineering from IBM.

Technical Skills: Python, C#, R, Flask, FastAPI, Reflex, WASM, Spin Framework, Docker, GitHub Actions, PostgreSQL, ChromaDB, LangChain, Hugging Face Transformers, BERT, XGBoost, VAR
Languages: Arabic (Native), French (Fluent), English (Fluent)

‍