This Atlanta Founder Has A Tool That Actually Understands Your Company's Custom Data

Atlanta entrepreneur Pavleen Thukral is ready to scale his next company.

Long-time Hypepotamus readers will know Thukral from his time building Stackfolio from 2015 to 2021. We followed the journey of his company brought on investors, won pitch competitions, grew its online marketplace for institutional loans, and brought on more than $1 billion in loan opportunities. The startup was ultimately acquired by Jack Henry & Associates in 2021.

With his latest startup, Magic Data, Thukral and his team are improving how companies work through their most complex data needs. At its core, Magic Data gives users “AI that understands custom data” and custom problems that come up from working across complex databases.

From startups to enterprise-level companies, Thukral saw time and time again that data engineering is often “the bottleneck,” since most organizations using a data warehouse have hyper-custom data schemas and the ingestion and transformations they do are difficult to maintain and scale, Thukral explained.

“Their data becomes difficult for external consumption (even by other teams within the same company),” he added. “Over time, as the complexity goes up, documentation becomes poor and cumbersome to maintain.”

The reality is that over 95% of data teams are working at or over capacity.

Data engineering can also be expensive, with the average data integration with a new source or destination costing $20K to $200K upfront.

Thukral saw an opportunity to bring a better, more custom solution to market with Magic Data.

The “Magic” Behind Magic Data

Magic Data got off the ground in February of 2024 following injection of outside capital from Christopher Klaus, famed Atlanta founder and investor. To date, Klaus is the only outside investor in the company. The Magic Data team got to work building a tool for custom data pipelines, and by November, Thukral felt confident they had a product and started recruiting for a private beta. Through the beta test, he realized that the pipeline product was important, but customers also needed more custom tools to help them focus on their data complexity challenges.

Magic Data is not “sector specific,” meaning that companies from across multiple industries can use the platform for complex data needs. But its recently launched SME product is an AI platform for data engineers working in a data warehouse environment.

The platform not only handles data from multiple sources, but can help ingest, transform, automate, and report customer’s most complex data needs. This is important because in our AI-driven business world, most of the big players like ChatGPT or copilot are not able to give specific and custom answers that data engineering teams are craving. And unlike other AI tools on the market, data engineers using Magic Data do not have to do a lot of context prep on their 100s of tables. They can just get to work.

Datasets and tables put into Magic Data can be understood at a semantic level, meaning it goes beyond basic keyword connections and can connect information based on overarching subjects and topics that might not be apparent when looking across multiple databases.

Thukral said that target customers are businesses with a high-level of data complexity (think hundreds or thousands of datasets and tables). He added it is particularly powerful for teams that are already using some sort of data storage and warehousing tools like Snowflake.

Lessons From The Startup World

Thukral is no stranger to Atlanta’s tech scene. Before founding Stackfolio, he founded HackGT (now HexLabs) while studying computer science at Georgia Tech.

As Thukral jumps into his next venture, he told Hypepotamus that he recognizes there are certain “privileges” that come from not being a first-time founder. Primary of these privileges is the fact that he has the ability to pull together a seasoned team of people who have worked with him in the past.

He also is bringing valuable lessons from his previous startup experience, saying that he learned the value of staying lean and getting creative about hiring the right talent, even if they join on a fractional basis.

--Photos provided by Magic Data