# Peeyush Agarwal

> ML Engineering @ Meta

Location: San Francisco Bay Area, United States
Profile: https://flows.cv/peeyush

## Work Experience
### Staff Software Engineer @ Meta
Jan 2023 – Present | Menlo Park, California, United States
Tech Lead for Fluent AI, Meta's AI Infra platform for e2e model development and publishing. Our SDK (FBLearner Fluent2 and Looper) is used by 1000+ engineers at Meta to build and deploy ML models to production.

### Staff Software Engineer @ Chime
Jan 2019 – Jan 2023 | San Francisco Bay Area
Chime, the leader in the U.S. challenger banking segment, helps members avoid bank fees, save money automatically, and lead healthier financial lives.

Lead SWE, ML Platform (2021 - 2023)
o Led ML Platform architecture, supporting team of ~15 eng.
o Built tooling to enable self-serve model development (DS experience) and deployment (batch and realtime orchestration).
o All of Chime's 20+ models developed, deployed & supported by our platform.

Founding ML Engineer (2019 - 2020)
Founding engineer of Data Science & Machine Learning team at Chime.
o Developed & deployed the first production ML model at Chime to automate mobile check reviews. ML model auto-accepts $xM of checks (xx% by count) from customers every day without agent
review at better-than-agent performance.
o Developed and shipped Feature Store to enable realtime models. Enables fetching 100s of features in
10ms latency and realtime inference in <200ms p95. Directly led to $xxM reduction in fraud losses
and ATO.
o Implemented tooling to enable Data Science team members to build and launch batch models.

### Software Engineer III @ Google
Jan 2018 – Jan 2019 | San Francisco Bay Area
Interpretable Machine Learning modeling for user task understanding and search re-engagement.
Developed, evaluated and productized unsupervised Machine Learning based models to identify user tasks (e.g. Car buying, Wedding Planning, etc.), label them and use them across various Google surfaces. Developed algorithms and workflows for running training and inference on massive (multiple petabytes) Google user activity data. The products powered by the models have received press converge from Google, Mashable, Engadget, AndroidPolice, CNet and many more.

### Software Engineer II @ Google
Jan 2016 – Jan 2018 | London, United Kingdom
Worked on Adaptive Brightness and Adaptive Battery features in Android P. Designed and implemented the Machine Learning algorithm, android framework APIs, and all the other infrastructure changes (like data pipelines, testing and evaluation framework) to enable this feature.

The features were launched in Android P. Adaptive brightness leads to ~20% reduction in slider interactions and Adaptive battery leads to ~30% reduction in CPU app wakeups. Both features were prominently highlighted in Google's Android P launch and received extensive press coverage.
https://blog.google/products/android/android-p/
https://youtu.be/ogfYd705cRs?t=2h19m

### Co-Founder @ Manaple Technologies
Jan 2013 – Jan 2016
Built a customer engagement platform for retailers. We acquired more than 40 paying customers from all over India including stores of brands like Reebok, W, Aurelia and Red Chief.

The platform enabled retailers to
*  Collect customer feedback and contact details on custom tablet screens at checkout counter
*  Send targeted SMS and email campaigns. Analytics & insights on collected data
*  Attendance and check-in for store staff

### Analyst @ American Express
Jan 2015 – Jan 2015
Developed ML models to drive higher card member engagement. Defined KPIs to capture performance on competing business needs. New model showed significant improvement to all KPIs. Designed and tested new features which led to significant performance improvements in existing as well as new models at American Express.

### Software Developer @ Google Summer of Code
Jan 2014 – Jan 2014
Designed and developed Data Analytics Log Manager, a platform for detailed user level data analytics. It provides
- An API for receiving log data consisting of arbitrary key value pairs
- Tools to filter, analyze and visualize this data
- Capability to create a parent-child relationship and add synthetic/computed data
- Authentication and authorization to help protect the data


## Education
### Master of Science (MS) in Computer Science
Stanford University

### Bachelor of Technology (BTech) in Computer Science
Indian Institute of Technology, Kanpur

### Secondary Education
St. Xavier's School, Jaipur


## Contact & Social
- LinkedIn: https://linkedin.com/in/apeeyush
- Portfolio: http://peeyushagarwal.com

---
Source: https://flows.cv/peeyush
JSON Resume: https://flows.cv/peeyush/resume.json
Last updated: 2026-04-12