In a typical organization, finance is one of the most important functions. But teams often get stuck in manual workflows. According to research from HR software provider Paylocity, 38% of finance teams spend more than a quarter of their time on manual tasks such as reviewing invoices.
Mathieu Hafemeister, a former fintech investor at Andreessen Horowitz, said he has seen many financial organizations struggle to scale due to manual tasks.
“The current state of the finance industry is a myriad of point solutions cobbled together within the finance department,” Hafemeister told TechCrunch. “Excel remains the lowest common denominator, with limited automation potential.”
Hafemeister points out that most finance departments actually rely heavily on spreadsheets. One study found that 82% still use Excel files for budgeting, forecasting, and other key financial planning activities.
After experiencing these frustrations firsthand while leading growth at fintech company Jeeves, Hafemeister teamed up with Ted Michaels, an old friend and former head of finance at Jeeves, to tackle financial tasks. I decided to launch a platform to automate this.
The platform, called Concourse, connects to a company's financial systems and allows finance teams to capture and analyze data, generate graphs, and ask ad hoc questions such as “What is our non-GAAP revenue?” I'll make it.
“Concourse enables finance teams to proactively uncover insights that help them be better prepared by enabling them to stay ahead of trends.” Hafemeister says Mr. “Instead of a tool trying to improve the speed or efficiency of completing a task, Concourse can be given individual tasks to perform entirely on its own.”
Concourse backend dashboard. Displays the status of various AI integrations and settings for fine-tuning them. Image credit: Concourse
Now, financial automation is not an entirely new technology. Linq recently used AI to come out of stealth and automate aspects of research for financial analysts. Ledge and Doopla also build a variety of finance-specific generative modeling tools.
But what sets Concourse apart is its ability to execute financial workflows that involve “complex, multi-step operations,” Hafemeister said. For example, the platform can pull data from a company's NetSuite dashboard, download a CSV file, and copy the data into an Excel spreadsheet.
“We leverage large-scale language models to perform optimal functions and combine them with more traditional data analysis techniques,” Hafemeister explained.
There is a lot of interest in AI in the financial field. According to one poll, 58% of finance teams are currently using some kind of AI technology, a 21% increase from 2023. Grand View Research estimates that the “AI in FinTech” segment, which was worth $9.45 billion three years ago, is growing 16.5% annually.
But to have any chance of impacting the financial automation technology market, Concourse needs to demonstrate the ROI of its product. This is a difficult feat. According to Gartner, demonstrating or estimating the value of AI is the biggest barrier to AI adoption for nearly half of companies.
Concourse also needs to ease potential customers' fears about errors and illusions caused by AI. A poll of UK-based executives by HR specialist Peninsula found that 40% said inaccuracies from AI tools were their main concern, followed by concerns about data confidentiality. .
Hafemeister said Concourse employs “a variety of tools and techniques” for fact-checking and verification, and strives to ensure that the AI performs its tasks as intended. He added that Concourse does not use company data to train its AI models, at least without explicit permission, and that the platform only collects data that customers share. .
“Data accuracy is paramount in the financial industry, and answers are typically either completely correct or completely incorrect,” Hafemeister said. “That's why at Concourse, we've spent a lot of time and effort delivering AI that can accurately perform assigned tasks. We also take data privacy and security very seriously and have partnered with the industry's best.” We built Concourse using practices.”
People seem happy to take Hafemeister's word for it.
Concourse is still in beta stage ahead of a wider release planned for next year, has multiple customers including Instabase and Shef, and is capitalized at $4.7 million. Hafemeister's former employer a16z is an investor in the startup along with Y Combinator, CRV, and Box Group.
Hafemeister said the focus right now is on product development and growing Concourse's New York-based staff of six.
“We raised funding to hire more engineers, build more AI-enabled workflows, expand our data integration reach, and begin expanding our go-to-market capabilities.” he said. “In engineering hiring, there is a focus on hiring backend, machine learning, and AI engineers.”