Portfolio
Selected Blog Posts
-
Compute-Saving Active Learning
-
Workflows for Data-Centric AI
-
Is Big Data Dragging Us Towards Another AI Winter?
-
MLOps as the Remedy to Tech Debt in Machine Learning
-
Did We Just Move on from MLOps without Anyone Noticing?
-
AI Sentience: Friend or Foe?
-
How We Got Responsible AI All Wrong
-
Impact, Bias and Sustainability in AI
-
Increasing Accessibility to AI
-
Creating More Opportunities in AI
-
What is Machine Teaching?
-
What is Data-Centric AI?
-
There's no Such Thing as "Green" AI
Selected Webcasts and Talks
DC Podcast, Feb 2021
My detailed views on how Active Learning will change the game in AI. I’ve dedicated a significant part of my career promoting Active Learning, and proposing ML-driven techniques to optimize an Active Learning process.
WiDS Keynote Panel, April 2018
One of many panels I have been part of to encourage the next generation of women to go into STEM and Machine Learning in particular.
The CUBE, 2018
My thoughts about the importance of diversity of thought in the context of innovation and the challenges associated with the creation of a healthy data culture within any organization.
ODSC West, Nov 2018
Active Learning was invented to address the raising cost of data labeling, but it really has applications in reducing the environmental impact of ML and in explainability. I cover some of those topics in this talk.
TWIML Podcast, June 2017
While at WalmartLabs, I pioneered the concept of MLOps, both as a technology and as a way to fight ML tech debt. This was years before the term MLOps eventually got widely adopted by the industry.
MLConf SF, Nov 2016
I believe the best leaders are versatile and lead by example. Remaining technologically relevant as a manager has been key to my career, and I remain a prolific inventor. This talk showcases a series of my inventions for the e-Commerce industry.
AI Infra Alliance, Feb 2023
DataPrepOps, a term I coined in January 2021, refers to the operational side of Data-Centric AI and Data Preparation of AI training data in general. Hear me explain it all on the topic.
Data Council SF, Apr 2017
In this 2017 talk, I discuss for the first time my concerns that the ML Community would eventually run out of training computational power.
My prediction is turning out to be accurate.
Data Council NY, Nov 2018
This is me evangelizing the concept of Active Learning…
Alectio Infomercial, 2021
This is me explaining the concept behind my company, Alectio, which is the first ever to focus on end-to-end DataPrepOps.
Berkeley Haas Podcast, 2018
Educating future executives on the topic of Data Strategy has been a recurring theme in my career. Accelerating the adoption of AI in the enterprise starts with Data Literacy…
ScaledML Panel, Feb 2020
An interesting conversion on the challenges of building AI products, and how those challenges relate to the structure of data teams.
Big Data Expo, Jul 2017
Human-in-the-Loop ML has long be one of my core expertise. This talk is one of the first I gave on the topic of the data annotation and HCI for AI.
TWIMLCon Panel, 2020
A deep conversation on how the cultural and organizational concerns to keep in mind to remove blockers to ML projects.
Berkeley Haas Podcast, 2018
A conversation with other women leaders in Data about the growing pains of the Data Community