Kevin AounDeep Learning Engineer
Skills
Technology | Softwares | Frameworks

Python | Pytorch | JAX | CUDA | C++ | LangGraph | Neo4j | NVIDIA TensorRT | NVIDIA Triton | AWS | Azure | GCP | Ollama

Professional Experience
inmind.ai, AI Engineer
  • •Helped clients right-size their AI infrastructure using NVIDIA tools, balancing performance needs against cost, often cutting projected hosting spend significantly
  • •Built cost models for complex AI deployments, giving leadership clear numbers to make decisions
    • Jun 2025 – presentRemote
  • •Took multiple AI projects from early prototypes to live production systems across Azure, AWS, and GCP, managed the technical tradeoffs at each stage
  • •Fine-tuned language models in-house rather than relying on external services, reducing per-query costs while maintaining accuracy for client-specific use cases
    • Lebanese American University & BMW Group, Research Assistant
  • •Researching automatic quantum state preparation using hybrid quantum-classical architectures, in collaboration with BMW Quantum Engineers and Physicists
    • Jan 2025 – present
    BMW Group, Internship⁠, DL & Automation Engineer
  • •Solved a long-standing data challenge by building a local LLM pipeline that processes 25k+ daily queries for data analysis, collaborated with Google and BMW stakeholders, now a flagship project
  • •Reduced latency by 85% and costs by 25% by optimizing the local runtime
    • Aug 2024 – Dec 2024Munich, Germany
  • •Presented the strategic approach and secured client approval for project expansion
    • Education
    Lebanese American University, BE, Mechatronics, Robotics and Automation Engineering

    CGPA: 3.8 | Dean's Distinguished List of the School of Engineering | Varsity Swimming

    Sep 2021 – Jun 2026Byblos, Lebanon
    Personal Projects
    HELIN, Humanity's Last Interface
  • •Building a brain-computer interface (BCI) for real-time motor imagery classification, epilepsy detection, and gaming
  • •Training CNN and Transformer-based deep learning models from scratch (EEGNet, LaBraM)
    • Sep 2025 – Present
  • •Curated high-quality datasets in collaboration with RIT Dubai, project featured on LAU website
    • GSE27, Graph Semantics Engine
  • •Turns your unstructured data (PDFs, URLs, YouTube..) into a structured graph for efficient retrieval
  • •Built on top of Neo4j, using ETL pipelines like Unstructured, Firecrawl, LlamaCloud
    • Awards
    Varsity's High Academic Athletic Achiever (AAA), LAU Athletics

    Issued to individuals with exceptional varsity season performance and top 3 academic standing

    2024
    1st place, Google Solution Challenge MENA Region, Google MENA

    Team project Khedne Maak ranked 1st among 120+ teams by Google experts in the MENA region.

    2022
    Honors Scholarship and Athletic Scholarship Recipient, LAU
    2021
    Extra Curricular
    President, Google Developer Students Club
  • •Hosted tech talks and speaker keynotes with engineers from Google ⁠
  • •Developed 5+ engaging workshop material delivered to 85+ university students
    • Aug 2022 – Aug 2024
    Certificates
    Certified SOLIDWORKS Associate (CSWA)
    Languages
    Arabic — C1English — C1French — C1German — B1
    Kevin Aoun
    1 / 1