FAQ

Absolutely. Many customers have already invested in external data, but find organizational appetite exceeds what their team can manage. Datastruct lets them take on more discovery projects, funnel their own sourced and developed datasets into our testing services, and push everything through our integration services.

Partnering with Datastruct lets you scale up or down without committing to permanent headcount, avoiding the risk of a long term hiring decision that may not match future demand.

This is common. Organizations tend to adapt their behavior to their organizational constraints. If you haven't invested in ongoing discovery and testing capabilities internally, it's no surprise consumption is low. The lack of investment also makes discovery and testing a risky proposition for most business leaders who will vastly prefer a clear path to success.

This is a case of induced demand: when that constraint is removed, appetite goes up. Business and data leaders start putting forward new use cases for external data, consumption increases, and meaningful business outcomes follow.

Yes, we support open data. Datastruct works across open, free, governmental, bureau, association, and commercial datasets, and we choose the best source for the customer's use case. Commercial data products can be valuable when they add years of engineering, cleaning, modeling, or structure on top of raw sources, but they are not always necessary.

When open data solves the problem, we will use it and integrate it. When commercial data is the better fit, we can support that too. Datastruct adds no margin on data costs, so we are not incentivized to push a commercial product over an open source. The incentive is to choose the product that works best for the customer.

While it would seem to follow that external data would be even harder to operationalize than internal data, we've seen the opposite time and time again. New external data feeds just don't have the constraints of internal data owners and systems.

It is not unusual for organizations to bring in new open or commercial data feeds in weeks to solve a critical business challenge, rather than spending quarters negotiating to modify or gain access to internal systems and data.

We work on yearly contracts that include a set number of discovery, testing, and integration projects, along with ongoing maintenance and support for integrated pipelines to Databricks, Snowflake, Salesforce, and API systems. Having prepaid resources opens up the organization to start using the service and growing in this area. We can scale plans to meet our customers' needs.

Datastruct works with some of the largest enterprises in the world, handling the most secure personal data elements those institutions support. We hold a SOC 2 and put data security and technology at the forefront of everything we do.

Our leaders have spent well over a decade working on discovery, testing, and integration with the largest banks and insurers. You can feel confident putting us forward to your information security, procurement, and legal teams.

Yes. It is not unusual for some data sets to have protected class or other sensitive elements that your legal or compliance teams are not comfortable having in your systems. Datastruct can remove these data elements from your integrations so that your systems meet your internal or regulatory controls.

Datastruct acts primarily as a processor in production, though we do resell data in production when needed. During discovery and testing, we regularly act as the contract holder and reseller. This lets an organization move quickly without needing legal involved before determining fit.

Datastruct handles all agreements with the providers. When there are fees, Datastruct adds no margin on top of data costs; those costs are typically pulled from a yearly data budget that's pre provisioned so your team can move fast.

Having a partner like Datastruct means you have the ability to switch to the best data whenever you want, with ease. Every year, we help transition data feeds from one vendor to another, ensuring no downtime and that key systems keep working as expected. This capability gives you leverage in negotiations with your data vendors that you didn't have before. While we're not offering cheaper data itself, we're enabling you to choose the best data at the best price, as needed.

Datastruct prices all of our integration work to be an obvious yes for our customers. We want it to be a clear return on investment. One way we prove that: any work we do to shape, rearrange, modify, or deliver data can be exported from our platform. You can leave Datastruct entirely, or move specific feeds out, whenever you want.

That may require some engineering work on your side depending on your frameworks and cloud providers, but you'll have all the logic you need, real code and tests, to port jobs quickly and effectively. We believe strongly in our offering, and we want you choosing us because we're consistently the best option, not because you feel locked in, and we'll keep working to make sure that's true.

Absolutely not. We work with some of the largest enterprises in the world, who have many data relationships outside our platform and technology. While the number of data sources we manage is growing, we don't recommend breaking or removing any existing data feeds to work with us.

In fact, we have features that help with your pre existing data flows, including an accounting feature that logs into your vendor portals, reads invoices, and gives you a standardized data model for data usage across all vendors, whether or not they run through Datastruct. This helps your accounting teams understand if they're paying the right amount for their data products.

While we're always glad to manage more of your data flows, there's real value even in partial use.

No. While we maintain a growing data catalog covering the world's open and commercial data, it's not like commercial marketplaces where you select a data product and buy it off the shelf.

We don't believe marketplaces solve the real problems data teams face with external data. The real problem isn't getting data into Snowflake or having a click to buy button. It's having the time and energy to do thorough discovery and vendor communication to narrow down the right set of sources, and then having the time, resources, and technology to run effective data tests to determine true fit for purpose. Marketplaces skip over these problems because they're hard. This is where Datastruct shines, and why so many customers have worked with us for years.

Both. We have technology that enables proactive discovery across hundreds of use cases our customers care about, along with specific technology that makes testing faster and higher quality. We also run a dedicated data platform that feeds APIs, warehouses, files, and other systems every minute of the day. Alongside that, we have a dedicated team supporting your team, your vendor relationships, and our platform and technology every step of the way.

Yes, absolutely. We offer both API and MCP integrations so you can request specific data attributes and get context friendly payloads that answer your agent's questions quickly and authoritatively. If your agent just needs to know that a specific property address exists, that can probably be handled by its training data.

But if you need to know that property's current market value, or whether its floodplain status has changed in the last nine months, you need up to date, trusted data delivered at the moment your agent needs it. We're seeing more and more use cases where Datastruct powers AI agents directly, and it's an area we're genuinely excited about.

No. Datastruct incorporates these requirements across our discovery, testing, and integration phases. During testing and analysis, we run waterfall analysis to determine how attributes flow given your specific input sets.

When we move to production, we use those same discovered algorithms and data flows to provision custom APIs or data feeds that deliver enriched attributes exactly as you need them. And because we act as an interface layer between you and the data vendors, if you ever need to swap one vendor for another, our interface layer supports that transition.

No. When we provide a custom API feed, we create a set of custom attributes in the payload for you to build applications around, something as simple as an email address or postcode. Behind the scenes, multiple data vendors may be filling out those attributes, and their full payloads are also included in the transaction, so you can capture that data in your warehouse through the API transaction itself.

We also offer asynchronous payloads of those API transactions, giving you the full data payload from what you've purchased. This protects your applications from being hard coded to specific vendor attributes, while also letting you retain all that data for future modeling purposes.

Yes. We work with a top five global insurer and reinsurer that has solutions all over the globe. We serve out of multiple AWS availability zones to reduce latency, and can solve your API or file based data challenges wherever they are in the world.

Yes. We have deep integrations with both Snowflake and Databricks, including data shares and materialized tables, as well as integrations with Salesforce and other data hungry systems. We can work with the technology your teams are already using today.

When Datastruct is helping a customer with real budget and a real production use case run discovery, we might reach out to anywhere from ten to twenty different data vendors to collect metadata that helps the customer determine fit for their use case. If that information isn't published on your website, we may reach out and ask you directly.

We can't promise any probability of being selected, but data vendors who provide the best, richest, and most accurate information are always more likely to advance to the next stage.

Absolutely not. Most of our production integrations are structured as processor agreements, where the data vendor has a direct relationship with our customer. What we're doing during discovery and testing is streamlining our customer's ability to explore what's available and determine market fit, moving to production as quickly as possible, for both the customer's benefit and the vendor's.

Our technology lets us move and work with data faster than vendors often can on their own, and our pre provisioned data budgets cut out what can sometimes be weeks or months of searching for budget to fund a data test. Our goal is to streamline the process, never to cut a data vendor out of the relationship. Often during testing, the vendor is brought in directly to talk with the customer and build their own dedicated relationship.

We want this to be an amazing sales channel for you: real use cases, committed resources to test and evaluate data, and a genuine shot at productionalizing the solution and developing an ongoing customer relationship.

Samples work for some of our customers, but often our customers have more sophisticated needs and a specific population they need to test against. We strongly prefer free data tests and do everything we can to minimize the work required on the vendor's side, including handling the actual enrichment through your APIs ourselves, so you don't have to lift a finger. That said, we understand data tests sometimes carry real costs, and we're fully capable of covering those through our data budgets.

Yes, and Datastruct uses AI where it makes discovery, testing, and integration faster or more cost effective. It is a useful layer for research, metadata extraction, use case framing, documentation, and accelerating repetitive analysis.

AI alone does not solve the whole external data workflow. Teams still need accurate vendor metadata, permission to use the data, real testing against representative populations, commercial and legal coordination, security review, production integration, and ongoing monitoring when a source changes. Those are the parts Datastruct is built to handle. We use AI to make the work more efficient, but the value is the combination of technology, vendor knowledge, testing infrastructure, and operational follow-through.