Machine Learning Researcher specializing in computer vision, multimodal systems, and scalable MLOps. Currently pursuing an M.Sc. in Machine Learning at KTH Royal Institute of Technology. I am dedicated to bridging the gap between complex AI research and robust, deployable pipelines, with a strong emphasis on multi-agent systems and spatial reasoning for autonomous agents.
- Multi-Agent Systems & Path Planning: Developing hierarchical planning frameworks and simulating autonomous agent logic (A*, Pure Pursuit) using C# and Unity.
- Generative AI & MLOps: Architecting RAG pipelines using LangChain, MLflow, and ChromaDB for semantic retrieval of unstructured telemetry logs.
- Aerospace Simulation: Engineering Hardware-in-the-Loop (HIL) systems for rigorous flight software validation with AESIR (KTH Rocketry).
- Languages: Python, MATLAB, SQL
- AI & Deep Learning: PyTorch, YOLO, U-Net, Hugging Face, Explainable AI (Grad-CAM, SHAP), LangChain (RAG)
- Computer Vision & Audio: EfficientNet, Swin Transformers, Multi-organ segmentation, Audio-Visual Fusion
- MLOps & Cloud: Google Cloud Platform (GCP), MLflow, ChromaDB, Git/GitHub, Linux
- Domains: Clinical Data Standards (DICOM, FHIR, OMOP), Multi-Agent Pathfinding, Systems Validation
- LungMate (AI-Driven Lung Cancer Diagnosis): Developed a hybrid CV system utilizing EfficientNet-B3, Swin Transformers, and YOLOv11 for multi-metric tumor quantification from DICOM data. Integrated a PyQt GUI for real-time clinical enhancement.
- AI-Driven Protocol Diagnostic Suite: Architecting a Generative AI pipeline leveraging ChromaDB and LangChain for efficient vector storage and semantic log retrieval.
- Radar Degradation & Beamforming: Engineered phased array antennas and real-time GUIs at DRDO and ISRO, achieving significant SNR improvements (>20%) and efficiency gains using beamforming algorithms (MVDR, LCMV).
- High-Altitude & Industrial IoT Safety Systems: Designed and applied for patents on real-time ESP32 embedded systems leveraging Sensor Fusion and LoRaWAN for critical safety telemetry in hazardous environments.
- Patents: Filed 2 Patents in IoT Safety Systems (Smart Helmet, Fall Protection).
- Publications: 15+ papers published in IEEE, Springer, and Elsevier.
- Awards: 2x Raman Award winner; LLM Evaluation Participant at Google DeepMind & Andon Labs.