Siyu Wu, Ph.D. Email & Phone Number

Washington, District Of Columbia, US

Managing a team of data scientists, ML engineers, and software engineers to deliver algorithms for next-gen ground-breaking hematology platform.

Washington, District Of Columbia, US

Washington, District Of Columbia, US

ML development for next generation hematology imaging machine.

Washington, District Of Columbia, US

Created a "first of its kind" ow cytometry gating ML algorithm for Leukemia with non-mutually exclusive cell populations. It reduced diagnosis runtime by 20 times while maintaining diagnosis accuracy requirements. Completed clinical trial successfully with a patent (as co-inventor) drafted. Product released on 9/22/2022.

San Ramon, California, US

• Leading enterprise ML product development, while wearing additional hats in UI, marketing, and sale.
• Exelon storm preparedness product suite-- Directing a team of data scientists, engineers, UI and UX designers to harden and scale PoV solution as a product for Exelon's 11M customers.-- Created and delivered end-to-end.
• Customer executive briefs and engagements-- Marketing and evangelizing ML solutions for customer executives.-- Worked with customers to accelerate change of management due to enterprise ML adoption.

San Ramon, California, US

• Had much fun innovating a "first of its kind" ML solution in aviation, and led a team effort for winning enterprise ML PoC competition in a much more matured market:
• N-step (n>10,000) ahead prediction of fatigue crack growth of aviation engine components-- Predict inclusion initiated crack growth (based on full flight data) accurately up to 10 years.
• Newmont predictive analytics challenge-- Led a team of 5 data scientists and innovated a ML solution that predicts ball mill failures up to 9 months prior to events.-- Won the competition against competitors.

Niskayuna, New York, US

• Repeated "Tiger Team" member for ML PoC competitions and analytical development across various verticals, and initiated and built infrastructure for multiple cross-departmental collaborations. Examples include:
• Remaining life prediction of helicopter engine components-- First FAA approved machine learning product (6/26/19), that reduced GE Aviation current physics-based practice from months to minutes, and is being.
• RWE power outage prediction-- Joined the project in the last week, turned around GE's initial bottom position, and won the best use case (out of 10) by outperforming competitors IBM and Teradata.
• Initiated and created R Shiny Server for fast UI prototyping and GE's first ML Blog for knowledge sharing-- R Shiny Server served as an indispensable resource for sales and marketing team for new customer engagement.

Høvik, Oslo, NO

As the key individual contributor, led the technical expansion of this econometrics based company from statistical methods to ML based approaches. Example projects include electricity load micro-forecasting for 4 million PG&E customers, impact evaluation and driver identification of energy efficiency programs, etc.

Merced, California, US

With 1,800 sensors that generate 1.2 million data points per day, my data-driven research focuses on improving building energy efficiency through fault diagnostics, predictive modeling of energy consumption, and control design and optimization. Results include further reducing electricity consumption by 5% of a highly energy-efficient LEED Gold building.

Beijing, CN

Designed and 3D printed the helix screen of the world's first 1500 mm Volumetric 3D Display.

Ph.D., Mechanical Engineering

M.S., B.S., Materials Science And Engineering