Mahmoud Bengounia

Developed a Retrieval-Augmented Generation (RAG) platform enabling clients to search contracts and PDFs efficiently by combining embeddings, vector databases (Weaviate/Qdrant), and BM25 algorithm for accurate results. Deployed a virtual assistant powered by Large Language Models (LangChain, DeepSeek, OpenAI, Mistral) to answer business questions directly from company documents. Integrated a robust AI stack including Hugging Face embeddings, FastAPI, and Docker to ensure scalability, reliability, and smooth production deployment. Created specialized AI agents to handle tasks such as data enrichment, anomaly detection, and workflow automation, streamlining operations.

Developed a RAG platform allowing clients to easily search contracts and PDFs, combining embeddings, vector databases (Weaviate / Qdrant), and BM25 for accurate results. Deployed a virtual assistant powered by LLMs (LangChain, Deep Seek/OpenAI/Mistral) to answer business questions directly from company documents. Integrated a robust AI stack (Hugging Face embeddings, FastAPI, Docker) to ensure scalability, reliability, and smooth production deployment. Created specialized AI agents to handle tasks like data enrichment, anomaly detection, and workflow automation, streamlining operations.

Automated workflows using Make and Zapier integrating Shopify/Stripe webhooks, REST APIs, Google Sheets/CRM synchronization, Slack notifications with retries, logs, and idempotency. Performed sales analysis and centralized data from e-commerce channels. Monitored B2B and B2C marketing campaigns across Google Ads, Meta Ads, Snap Ads, and TikTok Ads. Designed and automated reporting dashboards in Power BI to track key performance indicators such as revenue, campaign ROI, and acquisition costs. Conducted marketing automation and email marketing using Klaviyo.