Overview
Golden set out to organize information about companies, people and emerging technologies in a searchable knowledge graph. Founder Jude Gomila combined machine extraction with human editing to create connected records and citations. Golden later explored a blockchain-based protocol that would reward people for submitting and checking facts. In 2024, financial crime intelligence company ComplyAdvantage acquired Golden Recursion Inc. The acquisition applied Golden's data work to a more specific financial risk problem. [1][2][3]
Founding Story
Gomila had previously co-founded mobile advertising company Heyzap, which was sold in 2016. With Golden, founded in San Francisco in 2017 and publicly launched in 2019, he pursued a different problem: useful information about fast-moving technology was scattered across websites and databases. Wikipedia was a reference point, but Golden focused in part on subjects that traditional encyclopedias might not cover in detail. [1][4]
At launch Golden had raised $5 million. Its early contributors could edit pages, and the product linked claims to sources. The aim was to make research easier to inspect and update, while using software to help assemble information at a scale that manual editing alone would struggle to reach. [1][5]
Timeline
- 2017: Gomila founded Golden. [4]
- 2019: Golden publicly launched its knowledge base and announced $5 million in seed funding. [1]
- 2020: Andreessen Horowitz led a $14.5 million Series A. [6]
- 2022: Golden raised a $40 million Series B led by a16z crypto and described a token-incentivized knowledge protocol in testnet. [3]
- 2024: ComplyAdvantage announced its acquisition of Golden Recursion Inc. Gomila became a board observer and special adviser. [2]
What They Built
A knowledge graph represents entities and the relationships among them. For example, a company can be linked to its founders, investors and products. Golden's research engine let users search and filter those connected records, set alerts, and access data through a web interface or an API. Its documentation described human editing alongside automated collection. [5]
The product also sought finer-grained citations: a reader could see the evidence associated with a particular claim rather than only a list of sources at the end of a page. That made source quality and entity matching central to the product. If two people shared a name, the graph needed to avoid combining their facts. [1][5]
From late 2021, Golden explored a decentralized protocol. Its historical documentation describes contributors submitting and verifying facts, with token incentives proposed to encourage accurate work. In October 2022 the protocol was on a test network; reporting at the time described a planned 2023 mainnet launch. The cited sources do not establish that the proposed mainnet or token became a lasting production service. [3][7]
Market Position
Golden competed for researchers' time against Wikipedia, search engines and commercial company databases. Its proposed advantage was a connected, structured dataset with citations and an interface for querying relationships. Its early paid research engine focused on organizations tracking industries and could integrate with customer relationship management software. [1][5]
Business Model
Golden described its research engine as its first paid product. It offered search, alerts, research assistance and data access. For the later protocol, its documentation proposed free public access and charges for commercial use. These were distinct stages of the company's model; the sources do not establish their revenue or profitability. [5][7]
Traction
At the 2022 Series B, Golden said 35,000 individuals had participated in its testnet by submitting or validating information. This is a company-reported participation figure, not a count of paying customers. ComplyAdvantage later said Golden brought an extensive U.S. customer base but did not provide a customer count or acquisition price. [2][3]
Outcome and evolution
ComplyAdvantage announced the acquisition on April 24, 2024. It described Golden as a company automating a large knowledge graph and said its team would join the buyer's data science work. Gomila said he had known ComplyAdvantage founder Charlie Delingpole since 2005. Andreessen Horowitz also became a shareholder in the combined business. [2]
Golden's current website says its extraction and entity-disambiguation methods feed the buyer's data ingestion pipeline. That is a company account of the successor use, not a guarantee that the former independent research product or proposed token protocol continues unchanged. The transaction's price and Golden investors' returns were not disclosed in the acquisition announcement. [2][4]
Key Lessons
A large knowledge graph needs more than a way to collect facts: it needs sources, identity resolution and a reason for people to correct errors. Golden explored several ways to fund and organize that work, from a paid research engine to a proposed token system. The acquisition gave its data methods a specific application in financial crime intelligence; the public record does not establish why the broader original plan changed. [2][5][7]