Danny Goldman was up one night in 2021 working on a contract with a company that sells technology in parking garages.
As a former consultant to Bain and Co. working with private firms, he was responsible for collecting data from the 25 customer calls he took as part of his volunteering process. He spent five hours collecting customer data and still had to compile the first market analysis on the complete addressable market. At two o’clock in the morning, he found himself trying to find the Bain agency’s data on similar companies, but there was no way to conduct the research.
Although Goldman “totally attacked” that night, hitting a wall would give him the idea to start an idea that would become a reality.
On Tuesday, Goldman and his co-founder, Shivaal Roy, launched Mako AI after building the startup for nearly a year. Alongside the launch, Mako also announced its $1.55 million seed round led by Khosla Ventures.
The startup offers a Wall Street-facing AI agent that pulls a firm’s business data to answer questions and perform complex analyses. It aims to solve the problems of many former friends of the profession, such as collecting data, writing reports and analyzing companies.
“I was spending three to five hours a day looking for information, doing rote synthesis, building a formulaic proposal, and it was clear that something was broken here,” said Goldman, CEO of Mako. He added: “I started talking to a lot of people who had similar jobs, and it was clear that this is a universal truth.
Letting employees search an organization’s data has traditionally been tricky because of issues with user consent and data quality, said Roy, Mako’s CTO. But that hasn’t stopped some of the biggest PE firms from trying to improve corporate search. Blackstone, for example, built its DocaI platform for AI-driven search. At the same time, KKR developed RealHouse for its real estate teams to implement portfolio and data processing in one place for instant access to data.
How Mako works under the hood
Mako uses a variety of major languages, including OpenAI’s ChatGPT, to coordinate a network of multiple AI agents that can handle different aspects of a given task. Another important feature is a knowledge graph that tells models which documents are appropriate to answer certain questions. Mako runs this scenario of different documents during a 30-minute process, where different types of AI are used to read the documents and make sense of which ones are suitable for different types of information, Roy said. said.
If a user wanted to know what customers think about a given company, a knowledge graph might point to examples in the customer data-call, but if someone wanted to know more about revenue or the number of customers, that would be on the sound stage.
Prior to founding Mako and Goldman, Roy was the first engineering hire at Glean, a $2.2 billion AI-enabled business search firm. He said that some aspects of the Mako’s technology were built in the same way as the Glean.
Mako has also been trained in certain workflows common in private equity, such as benchmarking companies against each other, writing emails or drafting investment committee memos, Goldman said. Each result or statement, whether it’s a simple answer to a question or a full analysis, is cited in a separate document, which helps narrow down ideas, Roy added.
So far, Mako has several clients, including middle-market PE firms and growth-stage firms, including LA-based Shamrock Capital and San Francisco-based firm G.roundForce Capital. Mako, with one full-time engineer, plans to build their engineering organization through fundraising.
“The most important thing right now is to get this product from a first-year associate to a second-year associate,” Goldman said.
This is the site Mako used to raise its $1.55 million seed round led by Khosla Ventures.