Faizan Munsaf

Senior AI/ML Engineer

Devfied • Temps plein

Oct 2025 - Present • 1 yr 1 mo

• Architected and deployed a multi-agent RAG ecosystem on Azure AI Foundry, achieving sub-second response latency. • Engineered production Computer Vision pipelines, reducing false-positive rates by 35% across data streams. • Developed multi-channel workflows integrating WhatsApp, Email, and Discord for autonomous business actions. • Designed low-code automation layers, saving over 15 hours per week in manual overhead. • Developed an autonomous Agentic System for real-time meteorological data processing. • Built an end-to-end sentiment and emotion pipeline for dynamic TTS synthesis. • Integrated advanced facial recognition and vehicle tracking systems achieving over 94% accuracy. • Modernized core web applications using React.js and GraphQL, improving page load speed by 40%.

Senior AI/ML Engineer and Backend Developer

TransData International • Temps plein

Oct 2024 - Oct 2025 • 1 yr

Tech Stack: Python, FastAPI, Generative AI, NLP, RAG ● Led the development of scalable backend services using FastAPI and Django Ninja, applying the MVVT design pattern to ensure modularity and clean code separation. ● Built robust microservices using Docker, handling backend services, database migrations, reusable CLI commands, and messaging systems with RabbitMQ to support efficient CI/CD workflows. ● Integrated advanced NLP and Generative AI solutions using tools such as LlamaIndex, LangGraph, LangChain, and custom prompt engineering for intelligent document indexing, semantic search, and agent-based task automation. ● Contributed to cloud infrastructure and deployment across AWS, Azure (including Azure AI Foundry), and Heroku, ensuring scalability, high availability, and optimized resource management. ● Collaborated cross-functionally to improve backend performance, streamline Git workflows, and deliver scalable, maintainable systems. ● Worked on a GenAI-driven 3D avatar product, implementing lip sync capabilities and integrating Speech-to-Text (SST) and Text-to-Speech (TTS) models to generate lifelike, interactive avatars.

Senior Machine Learning and Backend Engineer

Software Alliance • Temps plein

Sep 2022 - Oct 2024 • 2 yrs 1 mo

Tech Stack: Python, JavaScript, TypeScript, FastAPI, Docker, AWS, Git, NLP, CV, GenAI, Prompt Engineering ● Led development efforts as a Python and private Blockchain Developer, delivering scalable backend solutions and AI-driven applications. ● Built and deployed ML/NLP pipelines for diverse use cases using techniques like beam search, keyword extraction, and custom RNNs. ● Developed Retrieval-Augmented Generation (RAG) systems utilizing LlamaIndex, LangChain, and AllenNLP. ● Fine-tuned Large Language Models (LLMs) with PEFT, QLoRA, AlphaQLoRA, and PPO for domain-specific applications. ● Designed and implemented robust backend architectures using FastAPI. ● Engaged in Prompt Engineering to develop tailored GenAI solutions. ● Fine-tuned YOLOv8 for object detection and tracking, optimized for real-time product detection (20+ FPS) over camera streams required because the person location changed and created many differences in a per second. ● Integrated TensorFlow Lite for mobile deployment, delivering a proof of concept (PoC) demo for client presentations.