Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi, I'm Mustafa Tariq, a Software Engineer and AI Specialist with hands-on experience building intelligent systems that integrate large language models, retrieval-augmented reasoning, and autonomous agent workflows. I design and deliver end-to-end AI solutions, including RAG-powered assistants, multi-agent pipelines, and production-ready inference APIs. I've built real-time computer vision pipelines, structured RAG chat systems, and custom LLM-based tools, and I thrive on turning complex ideas into practical, scalable software that users can rely on.…Hi, I'm Mustafa Tariq, a Software Engineer and AI Specialist with hands-on experience building intelligent systems that integrate large language models, retrieval-augmented reasoning, and autonomous agent workflows. I design and deliver end-to-end AI solutions, including RAG-powered assistants, multi-agent pipelines, and production-ready inference APIs. I've built real-time computer vision pipelines, structured RAG chat systems, and custom LLM-based tools, and I thrive on turning complex ideas into practical, scalable software that users can rely on.

Mustafa Tariq

AI Engineer, Full Stack Developer, Back-End Developer, +2





Hi, I'm Mustafa Tariq, a Software Engineer and AI Specialist with hands-on experience building intelligent systems that integrate large language models, retrieval-augmented reasoning, and autonomous agent workflows. I design and deliver end-to-end AI solutions, including RAG-powered assistants, multi-agent pipelines, and production-ready inference APIs. I've built real-time computer vision pipelines, structured RAG chat systems, and custom LLM-based tools, and I thrive on turning complex ideas into practical, scalable software that users can rely on.…Hi, I'm Mustafa Tariq, a Software Engineer and AI Specialist with hands-on experience building intelligent systems that integrate large language models, retrieval-augmented reasoning, and autonomous agent workflows. I design and deliver end-to-end AI solutions, including RAG-powered assistants, multi-agent pipelines, and production-ready inference APIs. I've built real-time computer vision pipelines, structured RAG chat systems, and custom LLM-based tools, and I thrive on turning complex ideas into practical, scalable software that users can rely on.

Available to hire

Hi, I’m Mustafa Tariq, a Software Engineer and AI Specialist with hands-on experience building intelligent systems that integrate large language models, retrieval-augmented reasoning, and autonomous agent workflows.

I design and deliver end-to-end AI solutions, including RAG-powered assistants, multi-agent pipelines, and production-ready inference APIs. I’ve built real-time computer vision pipelines, structured RAG chat systems, and custom LLM-based tools, and I thrive on turning complex ideas into practical, scalable software that users can rely on.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Language

English

Fluent

Work Experience

AI Engineer at Redback Operations

January 1, 2025 - January 1, 2025

Designed and deployed a real-time AI-based prototyping system using YOLO v3, enabling detection and storage of structured metadata (timestamps, frame IDs, head counts, bounding boxes) in MongoDB for time-series analytics. Implemented a distributed, event-driven architecture using Apache Kafka with a producer module and inference engines to support scalable monitoring pipelines.

Software Developer / Machine Learning Engineer at Red Marker Systems

January 1, 2023 - January 1, 2023

Developed and deployed ML-enabled software pipelines and inference services, enabling real-time detection and analytics; integrated with MongoDB for time-series data and leveraged Docker-based containerization.

Gen AI Engineer at Vedded

February 1, 2026 - Present

Designed and developed backend services for AI voice agents using FastAPI, exposing REST endpoints to support real-time workflows. Integrated external business platforms to automate tasks during live customer calls. Developed API-driven service layers connecting voice agents with Cliniko and ServiceM8 to retrieve and update customer, staff, and appointment information through their APIs. Deployed these services on AWS to ensure reliable and scalable access for voice agents during active conversations.

AI Engineer at Red Back Operations (Deakin Capstone Project)

July 1, 2024 - May 1, 2025

Designed and maintained real-time computer vision pipelines (YOLO-based) to detect faces, objects, and behavioral signals from video streams, persisting structured frame-level detections in MongoDB for downstream analytics. Built a retrieval-augmented generation (RAG) layer on top of computer vision outputs, transforming high-frequency detection logs into time-windowed semantic summaries for explainable querying and analysis. Implemented deterministic aggregation logic to merge frame-level events into meaningful time windows (10–30s), computing metrics such as average head count, anomaly frequency, and behavioral indicators. Embedded aggregated summaries into a vector database (ChromaDB) enabling semantic search and natural-language querying over vision data using LLMs. Enabled stakeholders to ask high-level questions while keeping answers grounded in raw detection data for traceability. Integrated event-driven processing with Kafka to support scalable ingestion and near real-time an

Software Developer / Machine Learning Engineer at Red Marker Systems

December 1, 2022 - May 1, 2023

Developed a real-time AI-based prototyping system using YOLO v3 and facial features to track face presence, head orientation, and directional movement, alongside detecting hands, pens, and mobile phones. Built a scalable MongoDB event-logging pipeline to store structured detection data for streamed monitoring, audits, and analytics. Deployed and optimized ML models on Azure, containerized with FastAPI, delivering low-latency inference and seamless integration via production-grade REST APIs. Improved automated monitoring accuracy by 40% through extensive testing, debugging, and cross-team collaboration.