Khalil Bezrati

– Designed and developed modular ETL pipelines using PySpark (locally), supporting full and daily incre- mental processing, organized in a medallion architecture. – Packaged and deployed PySpark workloads on Azure Databricks. – Automated cloud infrastructure provisioning on Azure (Databricks, storage accounts, and associated re- sources) using Terraform and ARM templates (Infrastructure as Code IaC). – Implemented CI/CD pipelines on Azure DevOps (trunk-based development) to automate infrastructure and data pipeline deployment, including orchestration of workloads via Databricks Jobs

– Contributed to the development of Inturn, an AI-powered platform connecting companies and interns while facilitating internship management. – Collaborated with backend, frontend, and UI/UX teams to gather requirements and translate needs into technical solutions. – Designed intelligent workflows for project planning and automated task assignment. – Developed backend APIs combining application data (PostgreSQL) with LLMs, producing structured JSON outputs directly consumable by the platform. – Supported the integration of AI features in the frontend, working closely with development and UI/UX teams for seamless adoption.

• Optimized the customer onboarding process by acheiving defined business requirement Leveraging AI models for various documents tasks, improving model performance and accuracy. • Automated classification and parsing of financial documents, increasing efficiency. • Designed and implemented prompts to guide multimodal AI models in detecting inconsistencies and potential fraud in documents.