Monitor ML Models
Deepchecks ML Monitoring

Monitor ML Models

Deepcheck ML Monitoring takes everything from the open
source packages and turns in into a secure, scalable
deployment that supports validating many models in
parallel.

Key Capabilities of ML Monitoring

Integrated Into Your Team’s Workflow

Integrated Into Your Team’s Workflow

Your team doesn’t work in a sterile environment, they want
Deepchecks to fit in with the tools they already have like:
Databases and blob storage for the logged production
data, PagerDuty, Slack and similar for alerting mechanisms,
and more. These are supported in Deepchecks ML
Monitoring.

Flexible Deployment Options

Flexible Deployment Options

Deepchecks ML Monitoring is built with various deployment
options, to take various data deployment needs into
account. Choose from On-Premises for complete control,
SaaS for convenience, or Single-Tenant SaaS with
dedicated resources for a mixture of the two.

Security and Identity Management

Security and Identity Management

Your data's safety is our top priority. We use Role-Based
Access Control (RBAC) to ensure only the right people have
access to the right information. Our Single Sign-On (SSO)
feature lets users access multiple services with just one set
of login credentials. This setting not only makes managing
access simpler, but also faster and safer.

Scalability

Scalability

The deployment of Deepchecks ML Monitoring is meant for
scale and reliability. It’s built to handle a large amount of
models in parallel, and works well even as the data scales.
While Deepchecks Open-Source is meant to contain
everything you need for a small scale, the deployment of
Deepchecks ML Monitoring is no better for production
environments.

Deepchecks ML Monitoring

Continuous validation of your data and models is vital for successful ML-based systems. ML and IT teams need to monitor
model performance, resolve issues quickly, and quickly find the root cause of production issues. Deepchecks ML
Monitoring provides flexibility, customization, and continuous usage, with robust infrastructure features build on top of the
open-source version.

Our Approach

Built with Context

Deepchecks ML Monitoring works in the context
of your ML, DevOps and IT solutions. We achieved
it by building our own system on top of a rich SDK
and a webhook for alerts. This enables you to
send data, configure monitors, alert rules and
trigger alerts in any 3rd party system.

Built for Scale

Our architecture uses scalable building blocks, which grow automatically as you scale the number of
models and amount of data you work with.

Built for Scale
Real Time

Real Time

When you have a production issue, you would like
to get a notification ASAP, which is exactly what
Deepchecks does for you. Deepchecks tracks
your model metrics in real time and raises alerts
as they happen

Secure

Deepchecks ML Monitoring meets your organization security needs, with the following security and
privacy table stakes:
Single Sign On

Single Sign On

Using “social login” (via
google) or your enterprise
SSO (Via SAML)

Data Privacy

Data Privacy

Keeping your data
encrypted on transit and
at rest

Data Separation

Data Separation

A multi-layered architecture
ensures that customers data
is kept separate

Secure SDK

Secure SDK

Gives you the capability to
extend our system in a
secure way

Open Source & Community

Deepchecks is committed to keeping the ML evaluation package open-source and community-focused.

Past Events

End-2-End Evaluation of RAG-Based Applications | LLM Evaluation
End-2-End Evaluation of RAG-Based Applications | LLM Evaluation
LLM Application Observability | Deepchecks Evaluation
LLM Application Observability | Deepchecks Evaluation
Config-Driven Development for LLMs: Versioning, Routing, & Evaluating LLMs
Config-Driven Development for LLMs: Versioning, Routing, & Evaluating LLMs

Recent Blog Posts

The Best 10 LLM Evaluation Tools in 2024
The Best 10 LLM Evaluation Tools in 2024