Resume

A compact overview of my AI engineering, computer vision, medical AI, and applied research background. You can also view & download my CV.

Profile

Alireza Hashemi

AI & Machine Learning Engineer with 4+ years of experience building production-ready systems across computer vision, medical AI, NLP/LLMs, and applied deep learning. I work across images, video, text, tabular data, and biosignals, with a focus on turning research-grade methods into reliable APIs, optimized inference pipelines, and usable products.

Core Skills

Programming & Data: Python, NumPy, Pandas, Scikit-learn, MLflow, Optuna

Deep Learning & Vision: PyTorch, Torchvision, OpenCV, Albumentations, Ultralytics YOLO, MediaPipe, MONAI, nnU-Net

NLP, LLMs & Multimodal AI: Hugging Face, Ollama, RAG systems, vision-language models, Florence, RexOmni, OCR/ASR workflows

Deployment & MLOps: FastAPI, Docker, Git, CI/CD, ONNX, TensorRT, quantization, GPU inference optimization

Research Interests: medical image analysis, neuroimaging, biomedical signal processing, brain-computer interfaces, responsible AI

Selected Projects

Multi-Task Vision Auto-Annotation Platform

Built a FastAPI/Docker system using Florence and RexOmni foundation models for detection, segmentation, keypoints, OCR, visual prompting, and scalable dataset generation.

Medical Imaging & Neuroimaging Pipelines

Developed breast cancer histopathology classification, brain tumor segmentation/classification, brain CT triage, and microscopy segmentation workflows using CNN, U-Net, transformer, and hybrid architectures.

Real-Time Vision Monitoring Systems

Implemented YOLO-based detection/tracking, people counting, face recognition, pose estimation, action recognition, and emotion analysis for applied monitoring and automation workflows.

LLM-RAG & Persian OCR Systems

Built domain-specific retrieval systems, Persian OCR services, and multimodal assistants that combine local LLMs, vector databases, OCR models, and image understanding tools.

EEG-Based Motor Imagery & Hand Movement Recognition

Designed CNN, RNN/LSTM, transformer, and spiking neural models for EEG classification and motor imagery decoding, supporting BCI-focused research.

Education

B.Sc. in Software Engineering & Technology

2019 - 2023

Mulla Sadra Technical Vocational University, Iran

Relevant Courses: Machine Learning, Data Structures & Algorithms, Databases, Python Programming

Certificates

  • Smartech AI Bootcamp: 120-hour intensive program in Python, ML, deep learning, computer vision, and NLP.
  • Deep Learning with PyTorch, Computer Vision with PyTorch, Generative AI Concepts, Machine Learning Fundamentals, Understanding AI.

Professional Experience

Senior AI Engineer

Jan 2025 - Present

Soorin Security

  • Developed AI-powered cybersecurity and asset intelligence platforms for network discovery, device fingerprinting, behavioral analytics, and SOC operations.
  • Built machine learning pipelines for network traffic analysis, clustering, anomaly detection, similarity search, and asset classification using large-scale telemetry data.
  • Designed and deployed production AI systems across Agentic AI, NLP, OCR, ASR, and LLM applications using FastAPI, Docker, and GPU-accelerated inference.
  • Applied representation learning and explainable AI to improve cyber asset visibility, threat investigation, and security intelligence workflows.

AI Engineer

Jan 2025 - Present

Hami System Sharif

  • Designed and deployed AI systems across computer vision, NLP, OCR, ASR, and LLM-RAG workflows.
  • Built a multimodal auto-annotation platform with Florence/RexOmni foundation models, supporting 20+ vision tasks through FastAPI services.
  • Delivered real-time vision modules for detection, tracking, counting, face recognition, action recognition, and emotion analysis.
  • Optimized model serving with Docker, ONNX/TensorRT-oriented deployment patterns, quantization, and GPU inference tuning.

Neuroimaging Researcher

Mar 2024 - 2025

SOAI

  • Researched EEG-based visual decoding, visual cortex modeling, and EEG-to-image reconstruction using spatiotemporal and generative approaches.
  • Developed reproducible BCI pipelines for motor imagery and hand movement classification using neural time-series models.
  • Contributed to research publications in motor imagery decoding and medical AI.

Computer Vision Engineer

Mar 2023 - 2024

Vida Medical

  • Designed medical image classification and segmentation pipelines for breast cancer, brain tumor, microscopy, and related biomedical imaging tasks.
  • Trained ResNeXt/attention-based histopathology models and U-Net/ViT-CNN style segmentation systems with rigorous evaluation protocols.
  • Prepared model APIs and optimized inference paths for clinical and research-facing workflows.

AI & ML Engineer Intern

Feb 2021 - 2022

Smartech

  • Built MediaPipe and Streamlit-based pose estimation systems for fitness and applied real-time vision projects.
  • Implemented YOLO/DeepSORT-style detection, tracking, counting, and emotion recognition pipelines for edge-oriented environments.
  • Improved training workflows through augmentation, preprocessing, and model evaluation routines.