Saad A S is a results-driven AI Engineer with several years of hands-on experience designing, building, and deploying end-to-end AI systems across Generative AI, Machine Learning, Deep Learning, and Computer Vision. He specializes in developing scalable backend architectures using Python, FastAPI, and Flask, and has strong expertise in containerized and distributed systems with Docker, Kubernetes, RabbitMQ, and Redis. Saad has delivered high-impact AI solutions including RAG-based and agentic AI systems, multi-agent…
Integrated Salesforce CRM for lead and company management with automated data exportation. Developed custom MCP clients for seamless multi-platform data integration and communication. Created intelligent multi-agent workflow that autonomously identifies qualified candidates by analyzing job descriptions.
Trained YOLOv11 model for accurate tile detection in real-time video processing. Implemented and optimized multiple Speech-to-Text models (Whisper, Parakeet) for audio transcription. Developed video processing workflows using FFmpeg and OpenCV for automated reel generation. Deployed scalable backend using FastAPI, containerized with Docker and orchestrated via Kubernetes.
Developed AI platform helping US citizens understand legal documents through conversational AI agent. Implemented automated legal document summarization and contract generation based on US laws using GPT-4. Created user-friendly interface with customizable prompts, generating professional PDF outputs.
Engineered pricing analysis system extracting product information from vendor PDFs using Gemini 1.5 and custom prompts. Implemented automated price and discount capture with price-to-value calculations across multiple vendors. Developed comparison analytics providing actionable insights on best vendor rates and deals.
Developed a computer vision system utilizing Roboflow inference for accurate detection of floor plan components including rooms, bathrooms, kitchens, and garages. Implemented detection algorithms for electrical symbols and furniture objects (chairs, tables, sofas) in architectural drawings. Enhanced image quality using OpenCV preprocessing techniques, improving detection accuracy by 25%. Created interactive floor plan modification features enabling users to add walls and detect natural language annotations on 2D floor plans.
Developed a privacy-preserving federated learning framework for detecting zero-day botnet attacks in IoT devices without centralizing sensitive data. Implemented client-server architecture enabling distributed model training across multiple IoT devices while maintaining data privacy. Evaluated system performance on NBaIoT and Bot-IoT datasets, achieving over 99% accuracy in botnet attack detection. Successfully defended thesis demonstrating the effectiveness of federated learning for IoT security applications.
Saad is available for hire
Schedule an interview14 days risk-free trial · No commitments · We handle contracts and payroll
Saad A S is a results-driven AI Engineer with several years of hands-on experience designing, building, and deploying end-to-end AI systems across Generative AI, Machine Learning, Deep Learning, and Computer Vision. He specializes in developing scalable backend architectures using Python, FastAPI, and Flask, and has strong expertise in containerized and distributed systems with Docker, Kubernetes, RabbitMQ, and Redis. Saad has delivered high-impact AI solutions including RAG-based and agentic AI systems, multi-agent…
Integrated Salesforce CRM for lead and company management with automated data exportation. Developed custom MCP clients for seamless multi-platform data integration and communication. Created intelligent multi-agent workflow that autonomously identifies qualified candidates by analyzing job descriptions.
Trained YOLOv11 model for accurate tile detection in real-time video processing. Implemented and optimized multiple Speech-to-Text models (Whisper, Parakeet) for audio transcription. Developed video processing workflows using FFmpeg and OpenCV for automated reel generation. Deployed scalable backend using FastAPI, containerized with Docker and orchestrated via Kubernetes.
Developed AI platform helping US citizens understand legal documents through conversational AI agent. Implemented automated legal document summarization and contract generation based on US laws using GPT-4. Created user-friendly interface with customizable prompts, generating professional PDF outputs.
Engineered pricing analysis system extracting product information from vendor PDFs using Gemini 1.5 and custom prompts. Implemented automated price and discount capture with price-to-value calculations across multiple vendors. Developed comparison analytics providing actionable insights on best vendor rates and deals.
Developed a computer vision system utilizing Roboflow inference for accurate detection of floor plan components including rooms, bathrooms, kitchens, and garages. Implemented detection algorithms for electrical symbols and furniture objects (chairs, tables, sofas) in architectural drawings. Enhanced image quality using OpenCV preprocessing techniques, improving detection accuracy by 25%. Created interactive floor plan modification features enabling users to add walls and detect natural language annotations on 2D floor plans.
Developed a privacy-preserving federated learning framework for detecting zero-day botnet attacks in IoT devices without centralizing sensitive data. Implemented client-server architecture enabling distributed model training across multiple IoT devices while maintaining data privacy. Evaluated system performance on NBaIoT and Bot-IoT datasets, achieving over 99% accuracy in botnet attack detection. Successfully defended thesis demonstrating the effectiveness of federated learning for IoT security applications.
Other vetted developers with similar skills and experience
