Reltio Connect

 View Only

Data Security, Data Integration, Data Quality: Your Questions Answered

By Chris Detzel posted 05-26-2022 08:48


In part one of our Ask Me Anything with Reltio founder and CTO Manish Sood, Manish discussed the motivation behind Reltio and the intersection of data governance and MDM. 

In part two of his interview, we will pick up where he left off on data quality, data security, and automating Reltio. Manish answers the communities questions.  

How should organizations fundamentally solve the data quality issues that emanate from poor source systems off record?

Manish Sood: I like to call it the method to the madness. There’s a sequence to approaching data. For example, say you don’t have an MDM or central 360-degree view of the information. In that case, all data is created and consumed in applications. Most of those applications don’t have good quality data quality capabilities or the ability to prevent duplicates from being created. In that case, the sequence that works starts by centralizing information instead of going across many different applications and managing the quality of each one separately.

Putting a system for this in place unifies and aggregates the information like an MPM system. You get more eyes on the data and understand the aggregate impact on the information. From there, you go back to the source systems to inform them where their processes are lacking.

The sequence centralizes the information to enable better insights into where you stand on the data quality paradigm. Then, with that centralized information, you can control the creation points in a manner where it is reducing the noise at those edges. The observability gets centralized, and the actionable metrics, and actionable, tangible capabilities, are tied to those applications.


All these activities take a lot of heavy lifting and manual operating to get right. What is Reltio’s to help move towards more automation or greater efficiency?

Manish Sood: That’s a great question. If you think about any deployments, whether Reltio or another MDM, there is a lot of manual effort even after the data is ingested, matched, and merged. You still have to resolve some manual matches or steward some data changes. And there are processes that customers how to define the anomalies and gaps in data. Can we automate those areas? Can we bring ML or AI algorithms to assist in that process? 

We created the Reltio foundation the way we did because we believe in the impact of AI and ML in this area. But to create a system where we could provide a recommendation that assists our users, we had to build the framework that would allow us to make those capabilities and ingest a reasonable amount of data from customers to start training some of the models that would aid users.

The great news is that we are now at the point where we are starting to double down on the AI and ML investment. A large part of the benefits that you will see out of it is things like detecting anomalies out of data, so you don’t have to fish for those anomalies. The system finds the irregularities and even suggests what fixes to apply. 

We’re investing in creating those capabilities that aid users. The first step is creating the capability of the data quality dashboards that we’re rolling out to early adopters. The next step is to identify anomalies that customers see from the application itself to reduce their efforts. You’ll soon start seeing improvements in those areas.

In matching and merging, there are lots of manual matches that the data stewardship team processes in the queue. In some cases, hundreds of thousands of manual matches need to be reviewed. In those scenarios, assistance from match IQ algorithms or models would help double verify the results so that the users can accept those results.

It’s just some of the examples that the team here at Reltio is very excited to unleash.

Reltio’s strategy is to attract applications and services to make their data and applications available through integrations with Reltio?

Manish Sood: Going back to something I mentioned, mastering is only 20% or less of the equation. Making data accessible across varied application landscapes, across every business process, analytics and data science type of outcome is a big part of the equation. One way to drive that is by making sure the integration options are easier and faster for our customers.

There are different ways we are going about that process. One is the rollout of the Reltio Integration Hub. The sole purpose was to have some API-level capabilities before the release of Reltio Integration Hub. It is great for developers who understand APIs. We had some connectors, MuleSoft, and SnapLogic-type integration toolsets that customers used with the Reltio connector. 

We’re all about making it easier. With the Reltio Integration Hub, we can bring the low-code and no-code types of flexibility into the mix. We have citizen developers developing integrations and taking data from Reltio to various applications or multiple applications into Reltio, so users have the closed-loop of integration. If one citizen developer creates the capability, they can share it with the rest of the community, so customers are helping to solve other customers’ problems as well.

We’ve created a platform, and we want to make use of that. In the past, we used third-party enrichment providers, such as Dun & Bradstreet. In life sciences, we have some sources that we put into a data-tenant type of orientation. We are creating a roadmap of different data domains and third-party data providers that offer pre0integrated capabilities. There are all different types of integration: some for enrichment, some for managing the entire lifecycle into Reltio, out of Reltio, into applications, and more. 

Our goal is to create a holistic set of integration capabilities so that more users can benefit from what already exists instead of building from scratch. 

Does Reltio have any plans to provide shareable data beyond Dun & Bradstreet?

Manish Sood: Several third-party data providers are relevant for the person, product, or organization domain. Our product management team is now looking at prioritizing some of those sources. We love to get feedback from our customers about which third-party data sources are relevant to them. Customers will inform a large part of our integration methods. Customers and partners are the best sources of helping us understand that landscape and tell us which sources are more relevant than the ones we have.

How does Reltio, a SaaS product company, assure customers with data security concerns?

Manish Sood: We are running a SaaS platform in the data management business where customers bring in their core critical business data. One of the parts of the responsibility that we have as a service provider to our customers is to ensure the security of that data. There is a significant investment at Reltio that builds up the compliance type of policies and invests in security technology that surrounds the perimeter of our product capabilities. And beyond the perimeter, too: they’re integral to how we create the product, how we engineer the code that rolls out into different environments, and all the checks that it goes through.

Security is a core part of our offering. It is one of our biggest investment areas after product capabilities. It’s a significant differentiator because we become a part of your security infrastructure. We provide security capabilities on users’ behalf, so they don’t have to create additional security around the data.

What the entire Ask Me Anything with @Manish Sood

Relevant Content: 
Data Governance, Master Data Management and How They Interplay
Ask Me Anything - Manish Sood - Founder and CTO - Ask here