TrueMLTalks #4 | Machine Learning @ Salesforce
TrueMLTalks #4 | Machine Learning @ Salesforce
TrueMLTalks is a video series in which we interview Machine Learning leaders and delve deeper into the ML stacks of their organizations. In this series, we will be speaking with ML-first organizations, like Gong, StichFix, SalesForce, Facebook, Simpl, and many more. We'll get a quick primer on their experience in managing ML pipelines and building best practices along the way.
Arpeet was part of the engineering team at Salesforce that built the entire ML platform. He is one of the founders of Builders Fund, where he and his colleagues invest and advise ML/AI companies worldwide. And at the same time, he is the head of infrastructure at Skiff.
The conversations with Arpeet covered the following aspects:
✅ ML Use Cases & Team Structure at Salesforce.
✅Overview of Salesforce ML Infrastructure.
✅Prototyping framework for the models going into production.
✅Managing Costs for Large-Scale ML Projects in the Cloud.
✅Managing the cost of infrastructure.
✅Building a Multi-Tenant Real-Time Prediction Service.
✅Optimization of models for enterprise AI.
✅Security and Reliability Measures in Salesforce AI Platform.
✅ML Infrastructure Platform vs Software Deployment Platform.
✅Advice for building infrastructure for ML operations.
Don't miss this opportunity to gain experience from one of the brightest minds in the ML industry. Tune in to this episode of TrueMLTalks now!
About Our Guest:
Arpeet Kale began as an intern at MetaMind while obtaining his master's degree in information science with a focus on data science. After MetaMind was acquired by Salesforce, he worked there as a member of the engineering team that built the entire ML platform at Salesforce. He is now the head of infrastructure at Skiff and one of the founders of Builders Fund, where he and his colleagues invest in ML and AI firms worldwide. Arpeet also assists several businesses on their ML and AI journeys.
⚡Reach out to Arpeet at the below link -
/ arpeetkale
About TrueFoundry:
TrueFoundry is a cross-cloud machine learning deployment PaaS that enables enterprises to expedite developer workflows for model testing and deployment while preserving full security and control for the Infra/DevSecOps team. We enable machine learning teams to deploy and monitor models in 15 minutes with 100% reliability and scalability, saving money and allowing models to be released to production faster, allowing genuine business value to be realized. We deploy on the customer's infrastructure, taking care of data privacy and other security concerns.
⚡ A short ~4 min demo of the platform: https://t.ly/0fk1
Read our blog: https://blog.truefoundry.com/
Subscribe to our newsletter: https://t.ly/oORp
#salesforce #automation #machinelearning
00:00 Start
03:52 The use cases of the AI platform at Salesforce.
06:47 The size of the team at Salesforce.
10:41 Tech stack on top of which the overall platform was built.
14:56 Prototyping the model in production.
16:41 Major challenges faced while hosting notebooks.
18:43 Managing the cost of infrastructure during the prototyping phase.
25:32 Building the serving layer that powered the models
37:10 Innovations that you and your team worked on at Salesforce.
38:10 ML infrastructure at Salesforce versus building it for a midsize company.
40:50 Changes in design decision you made before at Salesforce
43:54 Machine learning infrastructure platform versus a software deployment platform.
48:16 Investment into the ML/AI space and your views on the generative AI space.
52:07 Advice for the ML engineers or the ML infrastructure folks building infrastructure for ML operations in their companies.
Try our platform now - https://app.truefoundry.com/company-r...
Book a personalised demo - https://www.truefoundry.com/book-demo...
Watch video TrueMLTalks #4 | Machine Learning @ Salesforce online without registration, duration hours minute second in high quality. This video was added by user TrueFoundry 06 April 2023, don't forget to share it with your friends and acquaintances, it has been viewed on our site 407 once and liked it 12 people.