William Gervasio

Member of Technical Staff @ Microsoft AI

ResumeGitHubLinkedIn

Member of Technical Staff at Microsoft AI on Copilot, serving 100M+ monthly users and the youngest on the core product team. I own the release gate for all prompt and evaluation changes, coordinating five cross-team workstreams. Before that at Instacart, I built the receipt OCR pipeline processing ~86% of orders, improving recall from below 50% to 90% and saving $2–3M/month.

Previously built computer vision pipelines for cancer research at the Nobel laureate-founded Genome Sciences Centre and taught thousands of CS students as a Senior TA at UBC.

Member of Technical Staff - Core ProductMicrosoft
MicrosoftMicrosoft AIhttps://copilot.microsoft.com/
August 2025 - Present
  • Owned the release gate for all prompt and evaluation changes to Copilot's agent stack serving 100M+ monthly users — designing evaluation safeguards that coordinate 5 cross-team workstreams and prevent regressions in model behavior.
  • Co-architected the Copilot Tasks system prompt — Microsoft AI's first agentic execution system — developing prompt structures enabling long-running multi-tool agent behavior, including context trimming, action batching, and connector orchestration.
  • Designed and implemented backend integration architecture in C# enabling interoperability between two independent Copilot agent systems — auth flows, context routing, and event translation layers for cross-agent execution. Demo'd to Microsoft C-Suite.
  • Drove product retention improvements by designing and launching one of four statistically significant growth experiments in November — implementing backend logic, prompts, and evaluation framework in Python for long-term user engagement.
  • Established hiring standards for the organization's first early-career cohort by driving 300k+ social media views, sourcing 100+ candidates to interview rounds, calibrating the interview bar, and training interviewers.
MicrosoftCopilot Tasks: From Answers to Actions
CopilotMicrosoft Copilot
Software Engineer IIInstacart
InstacartInstacarthttps://instacart.com/
July 2024 - August 2025
  • Built the receipt OCR pipeline end-to-end — microservice, data pipelines, async jobs, algorithms, monitoring, and analyses — processing receipts for ~86% of Instacart orders. Improved product recognition recall from below 50% to 90%, saving $2–3M/month in operational losses.
  • Reduced end-to-end OCR training and inference pipeline latencies by 10× each, redesigning the real-time microservice with asynchrony and caching while optimizing offline model training from Pandas to native PySpark.
  • Developed a human-in-the-loop evaluation framework for ML model performance, enabling continuous measurement of precision/recall and guiding iterative model improvements across fraud, shopper quality, and item matching models.
  • Maintained and debugged critical card payment flows (Stripe, Marqeta) at 99.99% reliability SLAs. Implemented company-wide log sampling infrastructure for Python microservices.
  • Promoted to SE II ahead of the earliest standard timeline, ranking in the top 5% of performers within the first six months.
Instacart ShoppersInstacart Shoppers – Get Paid to Shop
Instacart AppInstacart – Grocery Delivery
Cancer Researcher, Machine LearningBC Cancer
BC CancerCanada's Michael Smith Genome Sciences Centre (BC Cancer)https://bcgsc.ca/
May 2023 - April 2024
  • Designed the end-to-end ML pipeline architecture for tracking 700+ cells dispensed per minute, linking physical and genetic cell features — a novel approach for a Nobel laureate-founded lab's cancer research.
  • Combined Bayesian inference, computer vision, and deep learning into a unified tracking system achieving 0.73 multi-object tracking accuracy.
  • Accelerated image labeling from months to minutes using motion-prompted foundation models on high-performance compute clusters, unblocking the broader research team to iterate faster.
  • Cut segmentation error rate by 50% with circular region proposals and data augmentation on Mask R-CNN (PyTorch).
Genome Sciences CentreCanada's Michael Smith Genome Sciences Centre
BC Cancer FoundationBC Cancer Foundation
Senior Teaching AssistantUBC
UBCUBC Computer Sciencehttps://www.cs.ubc.ca/
September 2020 - April 2024
  • Led programming labs and assessment across Software Engineering, Statistical Models, and Algorithms courses in R and Racket.
  • Designed and built the full autograding suite for Statistical Models, handling complex data outputs from ML algorithms.
  • Led delivery of UBC's Software Engineering MicroMasters on edX (188,700+ students); coordinated sprints, code reviews, and deadlines for full-stack projects in Racket and TypeScript.
  • Managed lecture and lab supervision, providing software design feedback and calibrating TA output quality.
  • Awarded Science Undergraduate Society Peer Helper Award for outstanding service and impact on peers.
CPSC 110CPSC 110 – Computation, Programs and Programming
CPSC 320CPSC 320 – Syllabus
Warrant Officer 1st Class: Fitness & SportsDND Canada
DND CanadaDepartment of National Defence - Cadets Canadahttps://www.canada.ca/en/department-national-defence/services/cadets-junior-canadian-rangers/cadets/summer-training.html
June 2018 - August 2020
  • Managed groups of 20+ cadets at a time while coordinating 3+ staff across barracks, drill, ceremonial, mental wellbeing, first aid, and morale. Coached cadets on how to lead others and conduct assessments.
  • Ranked in the top 13 nationally for leadership during regional staff placements; nominated best staff cadet in 2019.
  • Planned and led 18 weeks of sports and fitness training — organized tournaments, refereed competitions, and ran a tabloid event for 1,000+ cadets.
  • Commanded and coordinated 100 cadets on parade as Company Sergeant Major.
DND CanadaDepartment of National Defence - Cadets Canada
Vernon CTCVernon Cadet Training Centre
Feel free to send me a message.