BigPanda ML Engineering Meetup

On Wednesday, April 6th, Big Panda will be hosting a F2F meetup

18:00-18:30: Mingling & Pizza!
18:30-19:00: Lidiya Norman & Nadav Bar-Uryan - BigPanda - Building a modern DS work environment with Kubernetes
19:00-19:15: Short Break
19:15-19:45: Ori Peri - Riskified - Managing Multiple ML Models For Multiple Clients: Steps For Scaling Up
19:45-20:15: Or Itzary - Superwise.ai - Multi-tenancy isn’t just a glitch in the Matrix

***
Building a modern DS work environment with Kubernetes

Here in BigPanda we want our data science projects to be efficient, standardized and fast to deploy.
We envisioned an infrastructure that will allow our team to work with their prefered development methodology - scripts or notebooks, locally or cloud based while seamlessly transitioning between these options.
Our goal was to empower our DS with the capabilities they needed to increase velocity including local IDE, git, flexible resources, advanced debugging and a near production environment.

To solve this problem, we set out to create a simple and configurable infrastructure leveraging our existing Kubernetes capabilities that can be connected to our local IDE.

In this talk, we will share our journey from ideation to our current working solution.

Managing Multiple ML Models For Multiple Clients: Steps For Scaling Up

For most ML-based SaaS companies, the need to fulfill each customer’s KPI will usually be addressed by matching a dedicated model. Along with the benefits of optimizing the model’s performance, a model per customer solution carries a heavy production complexity with it. In this manner, incorporating up-to-date data as well as new features and capabilities as part of a model’s retraining process can become a major production bottleneck. In this talk, we will see how Riskified scaled up modeling operations based on MLOps ideas, and you will receive tools for how to set up your own continuous training ML pipeline.

Multi-tenancy isn’t just a glitch in the Matrix

Selecting, building, and maintaining the right ML multi-tenant architecture for your organisation while remaining sane.
Just like in software engineering, multi-tenancy is a choice, and challenge, of scale. Unlike software engineering ML multi-tenancy results in tenets with completely different model instances, features, hyperparameters, etc., running in production. Basically, the one model you started out with is about to take the red pill and go visit Alice.

What this session will cover:
Why model multi-tenancy
Multi-tenancy deployment architecture evaluation
Multi-tenancy observability and monitoring
What’s next? How deep does the rabbit hole go?