Use Orca In the Best Environment for Your Needs

Different teams have different needs, and Orca is built with flexibility. Simplified DevOps and fast set-up without having to worry about MLOps? Leverage Orca’s hosted SaaS version, where you can set-up and tune a model, then just call it via API for each inference.

Don’t want your data to leave your environment? Orca supports on-premise deployments where you use our solutions but supply your own compute.

SaaS 
(Serverless)
Orca provides all Orca products with:
Compute and storage provided by Orca
Model outputs provided via API access
Flexible, usage-based pricing

Set-up services:
Onboarding engagement to understand your use case
Guided memoryset tuning
Set-up SmartExplain agentic tools

Support:
Dedicated Slack channel with Orca engineering team
Live phone support
Quarterly optimization reviews
Optional professional services for hands-on help
Your environment 
(Cloud & On-Prem)
License to use Orca products:
Use your own environment (cloud or on-prem) to run the full Orca system
Simple set-up along with DevOps support for implementation
License based pricing

Set-up services:
Onboarding engagement to understand your use case and DevOps needs
Optional guided memoryset tuning
Optional custom fine-tuned embedding models

Support:
Dedicated Slack channel with Orca engineering team
Live phone support
Quarterly optimization reviews
Optional professional services for hands-on help

Start a pilot

Orca also offers flexibility to help you navigate practical challenges with running pilots. 

Some of our partners opt to leverage real datasets and complete each step themselves, with Orca’s coaching, to ensure the most accurate tests possible. Once successful, these pilots can also move quickly into production. 

Other partners, due to data sensitivity and other organizational concerns, elect to pursue more synthetic test cases and lean on Orca’s team to handle more work during this proof-of-value. These partners may also look to complete more complex pilots using actual data on the path to a full roll-out.

Thanks to Orca’s accelerated workflows and ability to replace synthetic memories with actual data without retraining, we’re happy to pursue the best structure for your team.

Getting Started

Orca, thanks to its ease of use and turnkey nature, can help you quickly set-up and assess how Orca solves your unique challenges.

1. Identify use case

We’ll work together to identify a use case, figure out what memory data to use, how Orca fits with your existing systems, and what Orca’s involvement in this pilot looks like.

2. Create starter memoryset

 You provide a sample dataset (either real or synthetic - it’s much more important for the data to have the proper structure) that can be converted into memories with Orca’s workflows.

3. Build & tune model

Once the memoryset is set up (usually within seconds or minutes),  use Orca’s agentic Data Management Co-pilot to optimize your starter memoryset and increase your model's accuracy.

4.  Run & adapt model

Begin running test scenarios (including live pilots) and leverage Orca's Inspector tools to continue refining your model's memory to drive outcomes

Find out if Orca is right for you

Speak to our engineering team to learn how we can help you unlock high performance agentic AI / LLM evaluation, real-time adaptive ML, and accelerated AI operations.