What you should know about the Tetra Integration and Tetra Data Libraries

October 30, 2023

Since TetraScience started building the Tetra Scientific Data CloudTM in 2019, we have made significant advances in how we plan, create, and deploy our data integration and data engineering components. Today, thanks to our customers’ support, TetraScience hosts the most sophisticated and fastest growing library of data engineering and data integration components in the industry. 

As we have progressed, it has become increasingly clear that providing clarity to our customers about how we build our project roadmap, and sharing best practices for component selection, is vital to our shared success. In this blog, TetraScience aims to outline our productization philosophy and component development processes to remove ambiguities and reduce friction in our collaboration. Below, we will answer the following questions: 

  1. What components are available in the library and what components will be added to the library?
  2. How can you request to accelerate development for items in TetraScience’s roadmap?  
  3. How can you request to add items to TetraScience’s roadmap?
  4. Why do we recommend biasing your implementation roadmaps toward existing, productized components?
  5. What to do if your vendor (instrument vendor or informatics application vendor) does not, or will not, support TetraScience and its efforts to build related integration and engineering components for your benefit? 

Components in the Library: Tetra Integration and Tetra Data 

  1. Tetra Integration: productized components that replatform data from heterogeneous sources of scientific data or publish data into customers’ applications/targets of choice. This maps to layer 1 on the scientific data journey pyramid—Data Replatforming. 
  2. Tetra Data (schema/taxonomy + transformation + documentation): we will call this “Tetra Data” for simplicity. These are data engineering products that transform raw scientific data into schematized, contextualized analytics and AI-ready data. This maps to layer 2 of the scientific data journey pyramid—Data Engineering. This is what customers typically refer to as an “IDS” (Intermediate Data Schema). 

Information provided to customers 

TetraScience publishes a list of all the general available (GA) Tetra Integrations and Tetra Data; and the roadmap of such components for the next 2 quarters. Every month TetraScience sends out a newsletter to our customers announcing new items and improvements made to the library. 

TetraScience philosophies 

TetraScience is dedicated to the continuous improvement of both our products and our services. We are also committed to ensuring the continued success of our customers. This is why we prioritize:

  • Component and component enhancement requests from existing customers compared with prospective customers.
  • Improvements and bug fixes to existing products compared with new products.
  • Work on generally available (GA) products that impact multiple customers compared with customized components that impact individual customers.

Prioritizing in this way helps us ensure product stability and quality.

When working with customers and prospective customers on new components, we will always focus on the highest value scientific use cases first. We believe that data integration and data engineering are simply means to an end. Our true goal is to accelerate and improve the scientific outcomes for your organization.

And please note, TetraScience also has training via both TetraU and our Tetra Catalysts packages to help your internal team deploy and use all the Tetra Integration and Tetra Data components. 

Plan your implementation with an understanding of our classification

Our team is excited to discuss how you can use TetraScience components and start planing your implementation. Before starting these discussions, it may be helpful to outline how TetraScience categorizes component requests.

To best serve all our customers, TetraScience classifies component requests into three categories: Productizable and generally available (GA), Productizable and can be accelerated, and Customized

Productized and GA: Immediate high value

We believe there are enough Tetra Integrations and/or Tetra Data Schema in our library for your organization to generate immediate value. These features are classified as Productized and GA

Items on this list include customer-agnostic Tetra Integration or Tetra Data components, which are deployable at multiple customers. Using productized components will not incur any Professional Services cost, except when you request that TetraScience set up, configure, and test these components. 

We suggest that you align your roadmap with our GA components as much as possible so you can achieve the economy of scale and immediate benefits supplied by productized features.  Again, here are links to our lists of GA Tetra Integrations and Tetra Data Schemas.

Productized and GA components are owned by TetraScience’s Product and Engineering Team. The Product and Engineering team also prioritize the roadmap by aggregating every customer’s input. If what you need is not on the list, your account team will bring the request to the Product and Engineering Team for prioritization. You can submit direct feature requests through Zendesk as well. 

However, we understand that the landscape of scientific data is extremely diverse and sometimes unpredictable. 

For example, a plate reader and its control software support a large number of export formats. TetraScience may choose to support specific report formats since they contain more comprehensive information and less ambiguity, thus making them suitable for integration with ELN/LIMS, analytics, and AI. In this case, TetraScience highly recommends you adopt the format TetraScience supports if applicable to your method/use cases. 

But, if you determine it is crucial to use your own file format, or your source system setup is outside Tetra’s support, then the request will be classified in one of the following two paths Productizeable and Accelerated or Customized.

Upcoming Product improvement: TetraScience is aware that it is currently not easy to find which Tetra Data (Schema/Transformation) is compatible with a particular vendor/model/type of instrument and documentation that describes the caveats and constraints of these components is sometimes lacking. TetraScience is actively remediating this gap, and is working on launching an internal facing tool called “Tetra Catalog” for the internal team to provide feedback. Our plan is to enlist early adopters from our customer pool to experience the Tetra Catalog in the first half of 2024. 

Productizable and can be accelerated: Influence Tetra’s road map with your use case

Customers can request that certain components, which are not yet available, be accelerated. Two types of components are eligible for acceleration. 

  • Roadmap items in the next 2 calendar quarters for Tetra Integrations and Tetra Data Schemas..\ 
  • Requests outside the roadmap, where TetraScience is convinced that this item can be productized and benefit multiple customers in the near future. Please note, multiple customers must make a documented request for this to be considered, or a significant part of the implementation must be reusable.

This category of features will be delivered by Data Engineers in the Professional Services (PS) team to minimize distraction and impact on the roadmap. Adherence to our roadmap’s timelines is vital to both our company culture and all our customers’ success. 

The TetraScience Professional Services (PS) team’s work will be reviewed by the Product/Engineering team. Engineering best practices will be followed to ensure quality. We will operate with the goal of eventually transferring ownership of this work to the Tetra Product and Engineering team for further improvements and maintenance.

All accelerated items will be added to the “common” namespace. 

All accelerated projects must be approved by the Product Steering Committee. The Account team will not be committing any work as accelerated until approved by the Product Steering Committee.

TetraScience will NOT charge you the full amount for acceleration. Items in this category have a reasonable chance of reuse and generalization. The exact discount will be determined case by case, and you will need to provide concrete scientific use case examples for justification. TetraScience will strive to provide fair pricing for such work.

Our goal is to operate with a productization/prioritization mindset. Since accelerated features are typically not field tested and carry a higher risk to implement, TetraScience would like to make sure this option is carefully considered to optimize customer success. 

 In exchange for discounted acceleration, TetraScience will typically ask for the following:  

  1. Customers should request timely support from scientific vendors (providing documentation/sandbox/SDK,etc.).
  2. A publicly endorsed case study post-implementation, to verify the scientific outcome of our mutual investment.

If your request can be fulfilled by something in the next quarter’s roadmap, TetraScience recommends that you wait for the release. In the meantime, you can ask for a preview of this item to ensure it fulfills your needs. We do not want you to wait for a quarter and then realize the release does not fulfill your requirements. 

Customized: Solve your specific data problem

At TetraScience, we understand that the scientific data landscape is extremely diverse. If you determine it is crucial to use your own file format, your source system setup is outside Tetra’s support, or you have a unique project you need help with, TetraScience is happy to design, create, and deploy a customized component, or help you to create those customized components in a self-service manner.

Customized component requests include items outside the product roadmap where TetraScience does not have sufficient evidence that the item is generalizable across customers.

Customization typically applies to the following scenarios:

  1. A new modality that is not commonly requested by customers, or a new file format, within existing instrument types. 
  2. Deployment, configuration, and testing of the product inside the customer’s environments or with the customer’s own instances of data sources/producers, data targets/consumers.
  3. Custom data validation or data processing pipelines.

Any customization will be delivered by Professional Services Data Engineers. Customized connectors, IDS, task scripts, and protocols will go into “client” or “private” namespaces if proprietary. If not, every artifact will be published into the “common” namespace. You can change and deploy anything published to the private namespace independently.

The Tetra Scientific Data Cloud also supports self-service pipelines where you or your team can create connectors, IDS, protocols, and task scripts to build out your own data integration and data engineering components. 

TetraScience also has training via both TetraU and Catalysts packages to help your internal team deploy and use all the Tetra Integration and Tetra Data components. The Tetra Scientific Data Cloud also supports self-service pipelines where you or your team can create integrations, data schema, protocols, and task scripts to build out your own data integration and data engineering components. 

Remaining the owner of your data

Unfortunately, some vendors attempt to, or inadvertently create a design, that restricts you from accessing your own data. Some of the most common restrictions include:

  1. Scientific data can only be read by the vendor’s software. There is no SDK/API or documentation on how to either export the data into non-proprietary formats or consume the scientific data programmatically. 
  2. The scientific data is contained in a file. File content has no predefined structure and has significant variations based on the instrument model, method, settings, etc. 

This is not a position we advise. TetraScience believes that your data belongs to you, as is customary in all other areas of the modern, data-centric B2B world we live in. Therefore, vendors have no say in how you can use and manage your own data. They must provide a non-proprietary means for you to consume your own scientific data. 

The Tetra Scientific Data Cloud is open and vendor-agnostic. We will work with any third-party scientific vendor of your choosing. We have no agenda other than replatforming and engineering your scientific data so you can move up the data pyramid and unlock analytics and AI capabilities. Ultimately, we are a scientific data cloud infused with our customer’s scientific data.

With or without TetraScience, you should be able to freely replaform and engineer your scientific data. If you can’t access your data because a vendor is holding it captive, please feel free to reach out and we will bring the appropriate resources to the table and help you prepare the right communication with your vendors. But, ultimately this is your responsibility to demand proper access to your scientific data. Once you have that access or knowledge, you can build your own integrations, data schema, data transformation using TetraScience’s self-service capabilities or work with TetraScience to influence and accelerate TetraScience’s roadmap. 

We look forward to working with you

We will schedule dedicated sessions during our next Quarterly Business Review (QBR) or regular meeting to provide voiceover, add context, answer your questions, and listen to your feedback. In the meanwhile, please respond with any questions or thoughts you have.