A lawyer and self-taught programmer with expertise across the full lifecycle of AI development: from building advanced NLP, speech, and computer vision models to fine-tuning LLMs and deploying them on scalable backend systems, often complemented by practical frontend development. Proven ability to deliver solutions for complex AI and data processing challenges.

Strong interdisciplinary background combining law, software engineering, and neuroscience-inspired problem solving. Experienced in applying signal-processing and adaptive modeling concepts to areas such as algorithmic trading systems and intelligent automation.

• PyTorch, TensorFlow, Keras, Hugging Face, Fastai, LAVIS, YOLO, Scikit-Learn, OpenCV

• Numpy, Pandas, Matplotlib

• FastAPI, Django, React, WordPress

• PostgreSQL, Pgvector, FAISS, ElasticSearch, MongoDB, InfluxDB, QuestDB

• Linux, Git, Docker, Kibana, Grafana

• Assembling GPU workstations

• Full Professional Proficiency

2017 – 2021

Dec 2025 – Mar 2026

May 2025 – Jul 2025

Sep 2023 – Dec 2024

Aug 2023 – Sep 2023

• Built a large-scale Persian question-answering dataset containing around 50k samples and approximately 200k lines, tailored for NLP and instruction tuning tasks.

• Fine-tuned Llama 3 on Persian datasets to improve Persian language understanding and generation capabilities.
• Worked with small quantized models and optimized them for Persian-language inference and generation tasks.

• Developed a retrieval-augmented generation system combining information retrieval and generative AI to improve accuracy and contextual relevance using external knowledge bases.

• Implemented a robust speech recognition pipeline optimized for noisy environments with high transcription accuracy.

• Designed and deployed a deep learning-based speaker identification and differentiation system.

• Developed a scalable image similarity search engine using embedding-based retrieval methods.

• Improved traditional face recognition workflows with optimized methods that significantly increased processing speed while maintaining accuracy on large-scale datasets.

• Performed unsupervised clustering of Amazon products using textual metadata, image embeddings, and co-purchase/co-view graph relationships.

• Developed custom scraping systems for structured large-scale data extraction and processing workflows.

• Implemented large-scale data management strategies using PostgreSQL, including indexing and query optimization techniques for scalable analytics and retrieval systems.

• Migrated search functionality from PostgreSQL to Elasticsearch across multiple NetBox products, improving scalability, search quality, and system performance.

• Implemented advanced resource management and optimization strategies to improve system efficiency, reliability, and operational stability under constrained environments.

• Engineered a stock market bot using web scraping and real-time financial data processing to generate actionable market insights.

• Built a trading web platform integrating backend and frontend systems for algorithmic trading workflows.
• Improved trading algorithm structures including TP/SL logic and trailing stops.
• Enabled fast backtesting pipelines and maintained infrastructure reliability.

• Developed an MVP LLM-powered chat platform with backend, frontend, and database integration.
• Implemented per-user chat history management and deployed GPU-hosted model instances using Docker.

• Built an end-to-end cognitive assessment web application implementing well-known neurological and cognitive tests.
• Designed and implemented backend systems, frontend interfaces, and UI workflows for complete testing pipelines.

• Dockerized and deployed production-ready systems with emphasis on scalability, reliability, observability, and maintainability.