Big Data

AHP: An Active Metadata Innovator – Atlan

Driving Social Impact with Democratized, Trustworthy Data

This series features Atlan customers who have recently completed a thorough evaluation of the Active Metadata Management market. Paying forward what you’ve learned to the next data leader is the true spirit of the Atlan community! So they’re here to share their hard-earned perspective on an evolving market, what makes up their modern data stack, innovative use cases for metadata, and more.

In this installment of the series, we meet Nick Murray, Senior Product Manager at Advocates for Human Potential (AHP), who shares how their team helps ensure every dollar spent by their clients has the highest-possible social impact, how a data catalog can act as a catalyst for the effective and efficient use of data, and the role technology plays in driving AHP’s data culture to new heights.

This interview has been edited for brevity and clarity.

Could you tell us a bit about yourself, your background, and what drew you to Data & Analytics?

I’m a Senior Product Manager at AHP, and the product owner of our “AHP Portal” application and our data governance platform, which Atlan fits into. My background is in UX design, particularly for data products. I have a background in the social sector, but I’ve spent time at large tech companies like Salesforce, at startups, and really found an opportunity here to do some impact work with some folks that I had met in the technology space years ago. So I was drawn to this company for the mission and for the position we’re in to up-level the way we help our clients use data for social impact.

Would you mind describing AHP and the role your team plays?

AHP is a consulting company and a contractor working mostly for federal and state agencies. Our service to them is technical assistance and training to help them decide who to fund with grant money, and then coaching those awardees through the use of those funds. That’s for things like opioid treatment and prevention. It includes things like the oversight and construction licensing of treatment centers, rehabs, and halfway houses. 

We’re a set of coaches that support the execution of substance abuse and mental health treatment on behalf of state agencies, but we also do infrastructure development. So usually it’s a different-agencies-but-similar-mission, the build-out and monitoring of facilities.

Our data team supports what we call gap analysis, or understanding where funding will be used most effectively, and we also assess pipeline/throughput analytics. For example, of the money our government clients are allocating and monitoring, where is it going and what is its impact? And then relative to that gap analysis, how far are our clients through a certain process of technical assistance or treatment strategy? The final piece is measuring outcomes as a result of that intervention. How has that gap analysis changed relative to a baseline?

Could you describe your data stack and processes?

We collect data primarily through direct primary collection – surveys and interviews with grantees and the grant applications they submit. We augment this with available open data sources that add context like socioeconomic indicators, health indicators, behavioral trends, and local government context. We try hard not to gather unnecessary data that clogs up the system without adding value but it’s a constant challenge to both serve the goals of multiple agencies without ever pulling our eyes off the objective of “IMPACT”.  

Amazon Redshift is our warehouse. It’s a lot of Amazon tooling. S3 is our data lake. Everything lands there before being ingested into Redshift. Then dbt is our favorite ELT solution for transforming that raw data into business use.

Why search for an Active Metadata Management solution? What was missing?

There’s a couple of reasons. The primary one that began the search was that we have a lot of users of our data that sit outside the data team. They’re doing their own analysis in Excel a little bit and BI tools, but we find that they’ll grab this data from different raw sources and just go off on their own and use it. We end up with multiple sources of truth and that leads people to cite the wrong details or make misinformed decisions, which we don’t want to see happen. 

Currently, a lot of effort goes into mapping where the truth lives across our organization. Ideally, it lives in our warehouse.

So then, the question becomes: “What are the means by which an individual who isn’t a data engineer would either access raw data they want to analyze themselves, understand the data behind a dashboard they’re using, or understand the data out there that’s available to them?” They need a catalog.

There’s another couple of reasons we sought out a solution on the technical side for the engineers themselves, just understanding and keeping track of everything going on in Redshift. But it was that external user use case that really drove us to search for these cataloging platforms.

Did anything stand out to you about Atlan during your evaluation process?

I was a big fan of Alation and so that was where we started. I ended up evaluating a few different platforms, and we did talk to our analyst at Gartner Research, who gave us a terrific overview of the current landscape. That conversation, plus our demo and sandboxing the various tools, led us to Atlan.

For me, Atlan stands out for its user experience. So if you think about the folks that we expect to use a catalog to get in touch with data, they’re non-technical folks, for the most part. So what is going to be the easiest thing for these folks to adopt and see the value in quickly?

You can imagine trying to sell a data catalog to a company like us. There were a lot of questions like, “Do we need this?” And it makes sense. You wouldn’t know. You wouldn’t know you really could make use of a smartphone until you had one. So user experience was key. The fact that it integrated obviously with our tooling, Redshift, dbt, PowerBI, those connectors were important too. 

But the other one was the attention we got from your sales team and technical team. So our warehouse is in GovCloud, which is complex by design. We had a lot of trouble standing it up on our own. Your folks were super helpful with that.

Landon Pangburn, the sales rep I was working with? I mean, he gave three people at my company his cell phone number. He was texting me on the weekend, “How’s it going?” That level of support was excellent, and I could tell, “Oh. That one’s a new company. You guys are interested in building strong relationships,” whereas folks with some of the older platforms are just hanging out, trying to figure out how to charge more money.

And then the pricing, given what we are looking at, you’re among the most competitive. That was a big factor too, just because it took a lot of selling on my part because data cataloging is not a thing that we do yet.

What do you intend on creating with Atlan? Do you have an idea of what use cases you’ll build, and the value you’ll drive?

In the near-term, it’s that more-or-less simple idea of, “This is where the source of truth lives. This is where our end users can go to find and understand the data behind the reports that we’re generating.” 

Similarly, in the near-term, we’re anticipating that our engineers use the catalog to document their own work so that they’re saving themselves some toil as they serve up these different reporting views. 

And we have a few teams, the major one we call the Project Teams. They’re the less-than-technical consumers of the BI layer. Then you have our engineers who are in the warehouse. In the middle are our data analysts, some of whom are new to things like database queries and views. Atlan is an educational tool for these folks to get more efficient at how they serve data requirements. The visual querying engine in Atlan, I think, leads very well into writing SQL. The fact that they can write test views and document them all within the catalog is huge for us. 

All this is very simple, basic stuff compared to what Alan’s capable of. What I’m interested in and what we’re interested in and have yet to really define, and we’ve been working with your customer success team to start this, is to identify, “Where else?” There’s a lot of other value here that we don’t even necessarily recognize yet. Like documenting automated tests with DBT-driven quality tests in our catalog, or auto-tagging sensitive information, which with our government clients is always a thing. 

Then there’s auto-tagging not just sensitive information, but information that’s coming from maybe less-than-verified sources. We collect a lot of information that we consider to be a source of truth. 

For instance, when we migrated to a new platform we used for previous rounds of funding, we needed to validate the data we were migrating from it. So when we put that data into our warehouse as the truth, we’d also like to flag it as, “Yeah, this came from the source” for the sake of clarity. And that’s just scratching the surface. 

And so from there and as we move towards employing data as more of a strategic asset and differentiating ourselves from other contractors or consulting companies in our ability to quickly wrangle and report on different data sources, we see our use of Atlan growing with us as we grow that capability.

Did we miss anything?

I think that one of the ways we’re equipped to have an impact in this day and age is with our use of data. Our sector lags behind the private sector in the use of solutions like this. So our mission as the tech team within this company is to bring folks like us to par with the private sector when it comes to this stuff. Because when we do that, we can have that much more of an impact.

Photo by Joel Durkee on Unsplash