Background

Dolt is a SQL database with Git-like version control features. We recently published a handful of case studies illustrating how companies are making use of Dolt in production settings. Of the companies that are using Dolt in production, some have required more than our free community support and have converted to initial paying customers. Providing them with support and professional service is a key part of our business model.

The Dolt database is free and open source, and always will be. We will continue to process bugs and feature requests from the community, regardless of whether they come from customers. The decision to become a customer should be based on the business value it creates.

Support

Contracted Dolt support came from the recognition that our customers using Dolt as a key part of their infrastructure needed more from us. Since commercial support is an enterprise standard for open source and non open source software and database tools, we had to figure out our offering. While we love being open source, we also have bills to pay. Our customers needing more and paying for it will allow us to keep investing in the Dolt project.

Our support model is currently a private shared Discord channel with each of our paying customers. We have a vibrant community around Dolt on our Discord server, and for those users that decide to become customers a private channel creates a dedicated space where we can offer more in depth support. This form of lightweight asynchronous communication is valuable, especially during the implementation phase, where long and expensive meetings reviewing lists of issues are made redundant by prompt resolution directly between engineering teams.

We offer two levels of support depending on the business needs of the customer.

Standard Support, 1 Business Day

Our standard support agreement is a good fit for use-cases like research pipelines. While these workflows can certainly be business critical, failure often does not warrant getting folks out of bed. Next day resolution normally suffices.

Turbine, a current customer, is a good example of this. They are building a simulated cancer drug discovery platform on Dolt, and are highly sophisticated. We have provided close support to them as they implement Dolt as their source of truth for the Simulated Cell model that is the basis for their simulations. This includes optimizing workflows to achieve better performance, providing guidance on best practices when working with Dolt, and finally all hands on deck efforts to fix bugs or performance issues that do occasionally arise. When those query performance issues do come up we have often been able to immediately provide a rewritten query as an immediate remedy, as well as identify the root cause and a longer term fix. Those fixes go straight back out to the Dolt community, a direct example of how paying customers enable us to keep investing in the open source project.

Enterprise Support, 24/7 Pager

For customers with more business critical needs around their Dolt implementation we offer Enterprise Support. In addition to the benefits of Standard Support, our enterprise customers get our CEO's cell phone number. If they suspect Dolt is causing their production issues, they can wake him up. As a former engineering director at AWS, and VP of Engineering at Snap, Tim appreciates the urgency of an outage.

Solutions Engineering

For customers that want to adopt Dolt, but lack the engineering resources or expertise to drive that adoption internally, we selectively engage in providing solutions engineering services. This involves partnering with our customers to design and deliver a data architecture around Dolt that creates the business value that motivated them to led to them discovering Dolt in the first place.

We are currently engaged in delivering a Dolt based infrastructure for a company that builds NLP models for their customers. NLP models are highly context specific, and a given customer may have many models for many contexts. Each of these contexts requires a separate training dataset that is being continually updated to reflect the evolving business needs and model performance metrics. Dolt provides a solution for marshaling those updates from analysts and engineers into the model delivered to the client, while providing full reproducibility and instant rollback across the model delivery pipeline.

Conclusion

If you have been evaluating Dolt but don't have the engineering resources to execute a project around it, we'd love to hear from you. We'd also love to hear from you if you are using Dolt in production and would like a support agreement with a specific SLA to support your business process. If neither of these situations are relevant, but you'd like to support the Dolt project directly you can do so via GitHub Sponsors program.

You can reach out via email, or come and hangout on Discord.