Clojure creator Rich Hickey has launched a new startup called Datomic, a spin-off from the consulting firm Relevance, Inc. Datanomic offers a cloud-based database of the same name that combines consistency and transactions with a flexible schema and scalability.

It’s an in-memory transactional database. Users get a set of tools to use locally, but the storage and processing takes place on Amazon Web Services. AWS’ recently launched DynamoDB is used to store all of the data.

Here are some of the features:

• Real-time updates
• ACID compliance
• Minimal schemas
• Immutable data (non-update-in-place)
• Its own declarative query language
• Data is stored in the cloud
• Distributed full text search

There are extensive technical discussions in the comments at Fogus and MyNoSQL, including many comments from Hickey addressing questions and concerns.

ServicesAngle

The needs for real-time analysis and less complex, more usable big data analytics tools continue to dominate the conversations, both online and at conferences like Strata. Datomic is an attempt to address both of those needs, and as you can see from the comments at Fogus, there are a number of potential benefits and trade-offs.

Although it’s coming out of a service provider, Datomic appears to be a very product-driven company. It’s coming down on the side of creating a simple to use big data product and minimizing the need for professional services to implement and maintain it. The need to store all data in the public cloud using DynamoDB may be a deal breaker for some, but it will also make deployment much easier for others.

Companies may find it’s more costly to adopt a complex open source product than it is to adopt a simple proprietary product. Lock-in can be a concern either way. The divide between big data companies with a services oriented business model and big data companies that have a proprietary product offering is an interesting trend to watch.