Customer Experience: How a Pharmaceutical Technology Company Manages the Accuracy of 100K+ Physician Records Daily

I recently sat down with one of our long-standing customers, a healthcare technology company based in Latin America that’s been with us since 2021. What started as a casual check-in turned into a fascinating conversation about the messy reality of managing physician data at scale, and why getting it right matters more than most people realize.

Here’s their story.

About Intelligo – A Pharmaceutical Technology Company

Intelligo, a pharmaceutical technology and consulting company, serves the pharmaceutical and consumer goods industries across Latin America. As a 100% Mexican company, they offer an integrated approach that combines technology, strategy, and information services, making them a strategic partner for clients who need comprehensive solutions without the complexity of managing multiple vendors.

At the core of their operations is a comprehensive physician database of over 180K doctors & physicians. A living resource that needs to stay current, accurate, and accessible. This database powers multiple technology solutions for the pharmaceutical industry, from medical sample distribution systems to promotional material management and healthcare provider engagement platforms. It is actively used every single day by pharma sales teams, marketing departments, and patients searching for care.

And here’s the thing about healthcare databases: they’re uniquely challenging.

A doctor might practice at three different hospitals, use variations of their name depending on the setting (E.g: Dr. María García vs. M. García-Lopez), change specialties, update credentials, or move locations. Multiply that complexity by 180,000 records, and you start to see the problem.

And that’s what Intelligo has to deal with day in, day out.

Key Challenge: The Duplicate Data Problem That Never Stops Growing

When I asked them about their biggest data challenge, the answer was immediate: duplicates.
“We’re constantly dealing with duplicate doctor records,” the Intelligo team explained. “And it’s not just about finding obvious duplicates—it’s about the edge cases. Records that match at 99% but aren’t quite the same person. Or records that ARE the same person but the system can’t tell because of how the information is formatted.“

This is the reality of healthcare data management. You’re pulling information from multiple sources—hospital registries, medical associations, pharmaceutical CRMs, public directories, insurance databases. Each source has its own format, its own standards (or lack thereof), and its own version of the truth. Dr. Juan Hernandez in one database might be Dr. J. Hernández in another and Juan A. Hernandez, MD in a third.

Before 2021, they were handling this manually. Can you imagine? A data manager and a technical team member sitting down to review records, make judgment calls, and merge information by hand. It was slow, it was tedious, and it was never really done—because new data kept flowing in daily.
The process took three weeks. Three weeks to clean, deduplicate, and prepare their database for use. By the time they finished, new duplicates had already crept in. It was like trying to empty the ocean with a bucket.

Preference: Why Intelligo Prefers On-Premises Data Deduplication Over Cloud & Modern AI Tools

When Intelligo started looking for a solution in 2021, they had one non-negotiable requirement: on-premises deployment.

When you’re managing physician data at this scale, you’re handling sensitive information. Doctor contact details, practice locations, specialties, affiliations—this isn’t data you can casually upload to a cloud platform and hope for the best.

“We have confidential data,” they told me. “We need to process 180,000 records daily, and we need complete control over where that data lives and how it’s accessed.“

Cloud-based data quality tools have their place, but for organizations dealing with healthcare information, the trade-off isn’t worth it. You’re introducing external dependencies—authentication servers, internet connectivity, vendor uptime, data residency questions. For a business whose entire value proposition depends on database accuracy and reliability, those dependencies become risks.
On-premises means control. It means the data never leaves their environment. It means processing happens locally, without latency or bandwidth constraints. It means they can work even when internet connections falter. And critically, it means they can meet their data handling obligations without relying on third-party infrastructure.

When I asked them about their experience with WinPure’s desktop platform, they said:

“It processes quickly and reliably, and there are no restrictions of any kind. There are no errors or unexpected shutdowns.“

For a team processing 180,000 records daily, reliability isn’t a nice-to-have. It’s the entire foundation of their operation.

The Daily Workflow: Cross-Referencing at Scale

Here’s what their daily data quality process looks like now:

Their data manager works alongside their technical team to run cross-reference operations across their physician database. They’re matching records from multiple sources, identifying duplicates, and cleaning inconsistent formatting—all through WinPure’s interface.

The features they use most? Database cross-reference and data cleaning. These aren’t fancy AI-powered predictions or black-box algorithms. They’re straightforward, deterministic matching operations that give the team visibility and control over what’s happening to their data.

“The environment is friendly and easy to use,” they told me. “The processing speed is fast, and there are fewer errors.”

But here’s where it gets interesting: they’ve had to fine-tune their matching thresholds over time. Setting a match threshold too low means you catch everything—including false positives that aren’t actually duplicates. Setting it too high means you miss legitimate duplicates that should be merged.

They mentioned their biggest challenge: “Avoiding false duplicates at 99% match confidence and having to modify parameters or increase the match percentage to get it right.”

This is the art of data matching at scale. It’s not just about running an algorithm and trusting the output. It’s about understanding your data well enough to know when the system needs adjustment. And having a platform that lets you make those adjustments without writing code or filing support tickets.

Results: From Three Weeks to Three Days

Let’s talk about the impact.

Remember that three-week process I mentioned earlier? The manual review, the judgment calls, the painstaking record-by-record analysis?

It now takes three days.
Not three weeks. Three days.

That’s an 85% reduction in processing time—time that can now be spent on higher-value work. Analyzing trends in physician data. Improving the UbicaDoc user experience. Supporting pharmaceutical clients with better targeting and distribution strategies.
But the efficiency gain isn’t just about speed. It’s about consistency. When you’re doing something manually, quality varies based on who’s doing it, how tired they are, what else is competing for attention. Automated matching with configurable rules means the same standards apply every time, to every record.

It also means they can actually keep up with the daily influx of new data. 180,000 records isn’t a static number—it’s a living database that needs continuous maintenance. Physicians change locations, update their practices, retire, join new hospitals. Without fast, reliable deduplication, the database would degrade in quality every single day.

The Bigger Picture: Pharma Data Quality at Scale

Intelligo’s case study illustrates something important about data quality work in sensitive, high-stakes industries like pharmaceutical companies: when your business model depends on accurate data, when inaccurate records mean wasted marketing spend, failed patient searches, or compliance risks—you can’t treat data quality as an afterthought.

Want to know how much money you could be losing to bad data? Find out using our cost calculator.

You need tools that:

• Give you control over where your data lives and how it’s processed
• Process at scale without degrading performance or requiring cloud infrastructure
• Provide visibility into matching logic so you can fine-tune for your specific use case
• Work reliably day after day, without errors or unexpected failures

For Intelligo, that tool is WinPure Clean & Match. For four years, it’s been the backbone of their physician database operations—turning a three-week manual process into a three-day automated workflow, while keeping 180,000 records of confidential healthcare data completely under their control.

And the fact that they keep renewing, year after year? That tells you everything you need to know about whether it’s working.