Matti Akbari
Technical Program Manager - he Engineering (Cesi) @微软
Signup · Get unlimited contacts
WORK HISTORY
Technical Program Manager - he Engineering (Cesi) @微软
Vancouver, BC, CA
BAU - Leading a team of 5 engineers to structure and onboard telemetry data for data center infrastructure Leading OpenAI data center telemetry onboarding, working closely with lease providers, site teams, tooling, and reservation teams.OpenAI DATA CENTER ONBOARDING - Delivered 110% of scope ahead of RFS milestones for GPU cluster telemetry, accelerating platform readiness and boosting GPU utilization across multiple high-priority sites.AI DOCUMENT PROCESSING PIPELINE - Architected an end-to-end AI pipeline (Azure OpenAI, YOLOv8, Python) that cut PDF-based hierarchy extraction from 5 days to <4 hrs, scaling data-prep for downstream teams.MODBUS MAPPING AUTOMATION - Engineered a solution that eliminated a critical workflow bottleneck, reducing manual engineering effort from 3 days to under 10 minutes (5 min creation + 5 min validation). This initiative is projected to save $1.5M annually across a team of 15 engineers and TPMs.
EDUCATION
Azad University (IAU)
Bachelor of Engineering - BE, Industrial Engineering
The University of British Columbia
Master's degree, Master of Engineering Leadership
ABOUT MATTI AKBARI
AboutAs an AI Tooling Architect and TPM, I build the intelligent systems that eliminate repetition and scale infrastructure performance. With a foundational background in embedded engineering, I bring a systems-first mindset to complex challenges in Robotics, Data Center Telemetry, and Industrial Automation. My focus is simple: deliver scalable, resilient solutions that generate measurable business impact.Selected Projects & Achievements (20••••25)OpenAI Data Center Onboarding & AccelerationSpearheaded the end-to-end onboarding (HE Engineering Team) for high-priority GPU clusters, delivering 110% of the initial scope ahead of all Ready-for-Service (RFS) milestones. My leadership in telemetry integration and system validation directly accelerated platform readiness and measurably improved GPU utilization across the AI stack.(BAU)AI Document Processing PipelineArchitected and deployed a full-stack AI pipeline with a team of 5 engineers in under two months, enabling the digitization of electrical devices from PDFs in a hierarchical format and slashing overall document processing time from 5 days to less than 4 hours. The system leverages a multimodal architecture (combining LLMs, Azure Document Intelligence (ADI), OCR, and Object Detection model training and data preparation), a Docker and Kubernetes-powered deployment stack, and a RAG (Retrieval Augmented Generation) implementation for superior LLM accuracy, performance, and scalabilityIndustrial Automation & Process OptimizationEngineered a Modbus mapping automation solution that eliminated a critical workflow bottleneck, reducing manual engineering effort from 3 days to under 10 minutes (5 min creation + 5 min validation). This initiative is projected to save $1.5M annually across a team of 15 engineers and TPMs.Microsoft Global Hackathon Winner (2024)Recognized for innovation and technical excellence, winning 3rd Place Overall and Best Accessibility Project. My team developed a computer vision–based accessibility tool on a robotics platform, utilizing Azure Custom Vision to create a real-world solution for users with disabilities.I’m now exploring leadership opportunities in Solution Architecture and AI consulting—let’s talk if you’re building anything at that intersection
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.