Overview

Riveter, a female-founded, Y Combinator-backed company, is redefining corporate strategy through deep market research and analysis. As a growth-stage startup, Riveter specializes in rapidly prototyping solutions for clients, often processing lengthy and complex financial documents like 10-Ks and M&A reports to deliver actionable insights.

The Challenge

Document complexity + Transformation volume

While scaling their product as a small team, processing a high-volume of documents (that are oftentimes hundreds of pages long) proved to be quite the challenge. Traditional tools struggled to handle this scale and complexity of data, which led to a great deals of errors and barriers for the Riveter team. "I think our largest current scale challenges revolve around the size of the documents and the number of documents we process at once" Erica Clark, CTO of Riveter, shared.

Transformation but no contextual understanding

Riveter's solution provides deep analysis of financial information for clients like Gusto, so processing and transforming data and providing a tabular output was not enough. A need for contextual understanding felt necessary but difficult. "What's the unit? What's the scale? How can we understand this from the documents we're processing? With a few tools we've used, we were able to turn documents into data (and pull numbers) but things would be represented as 2.6 for example when the real data was 2.6 million" she voiced, regarding their pursuit of the right solution.

The Solution: Implementing bem to scale data transformation

Riveter adopted bem's platform to support their tech stack (Typescript, Ruby on Rails, and Python), specifically tailored to extract and contextualize financial data from diverse document formats. "bem was really easy to pick up, get familiar with, and start using" Erica shared. bem has supported their journey so far with:

• Contextual Data Extraction: bem enables Riveter to both extract tables and numbers and to clarify units, scales, and other measures. Deeper insights can now be scalable- aka, we provide the output of 2.6 million.

• Rapid Prototyping: With bem, Riveter can swiftly prototype new data schemas and processing pipelines, reducing development time from days to mere hours. This agility proved invaluable during iterative refinements and pivots within their service offerings.

• Scalability and Flexibility: Despite handling diverse file formats and large document sizes, bem allows Riveter to scale their operations seamlessly, managing multiple pipelines and adapting quickly to evolving client needs.

Accuracy at Scale for Market Analysis

Since trusting bem with their financial modeling data, the Riveter team have been able to prototype complex data pipelines quickly for their analytics work, providing a deeply contextual experience for the inputs they're receiving. Faster production and improved accuracy allows their founding team to scale meaningfully for their clients.