Revolutionizing Data Management in Healthcare: Memed Modernizes their Data Architecture with Dremio

The Customer

Memed is an innovative Brazilian healthtech company transforming doctor and patient interactions through digital prescriptions. Memed integrates with providers and expands the traditional patient journey into a marketplace of drug stores, offering a more seamless and efficient healthcare experience. Not only do they simplify the patient's process, but they also provide valuable business intelligence solutions to all the players in the healthcare ecosystem, making processes better and offering solutions to hospitals, clinics, health insurance, and pharmaceutical companies. Memed's approach even improves the lives of doctors by providing a comprehensive view of all drugs and their clinical interactions.

The Challenge

Memed's revolutionary approach generates a vast amount of data, and managing this data became a significant challenge. To manage the growth of semistructured and unstructured data, they built a data lake solution. Their initial data lake was too slow and often unreliable, leading to missing data and frequent Extract, Transform, & Load (ETL) crashes. Instead of a proper, well-structured data lake that could provide a single source of truth, Memed was left with a faulty system that presented metrics in varied ways. Any changes to the architecture were cumbersome and prone to error, inducing collaborators to avoid working on the data lake and instead implementing the transformations straight within the dashboards.

The Solution

Memed, lead by Gabriel Rosario, Head of Data, initially analyzed multiple tools through a series of PoCs to not just overcome these challenges, but also to set up Memed's architecture for future scale and flexibility. With Dremio's ease of use, performance, openness, and ability to enable SQL transformations efficiently and effectively, the Memed team found their ideal partner. The new architecture with Dremio provided both the necessary structure and speed for Memed's data lake.

Dremio's solution enabled quick ETL processes and made use of Reflections, which are essentially materialized views that optimize data on read and are designed to automatically improve performance. It supported the data engineering team in managing complex tasks while allowing the analytics engineers to build the higher-level layers. Additionally, Dremio enabled the materialization of components outside of its environment and offered a single source of truth for all internal users.

The Results

The introduction of Dremio into Memed’s ecosystem led to significant improvements. They were able to complete a daily ETL process that previously took 30-40 minutes each in just 40 seconds, and the cost of their data stack was reduced by 10%. Even in the middle of a meeting, an analyst is able to run a full set of queries in just 10 minutes, providing all the necessary answers. Without Dremio, these tasks would have taken all day

The data team at Memed is more productive because of Dremio. They spend far less time waiting on query results, and they spend almost no time on ETL processes, so they are able to shift their focus from daily routine maintenance of the data stack to more strategic data
work.

With their previous stack, analysts were limited in the amount of transactional data they could query at a time. With Dremio’s acceleration and optimization capabilities, analysts were no longer limited in the amount of data they could query. Dremio also solved the problem of loading data, ensuring that all dashboards were accurately updated and interconnected. And it wasn’t just an improvement for their internal users. “As an integral part of the healthcare ecosystem, we have many external clients, including doctors, patients, hospitals, and drug stores,” said Gabriel Rosario, Head of Data at Memed. “These clients rely on the data we collect and deliver to give and receive effective care. With Dremio as our data lakehouse solution, external clients who rely on our dashboards now have access to accurate and timely insights.”