In legacy industries like financial services, insurance, and lending, analysts still spend hours every day buried in documents, triaging emails, retyping information from PDFs, and manually populating CRMs. For Heron Data, a New York startup founded in 2020, this market inefficiency was an opportunity.
Heron says they now process over 350,000 documents a week for more than 150 companies, including FDIC-insured lenders and national insurers. The company recently announced it had raised $16.6 million in Series A funding led by Insight Partners, with participation from Y Combinator, BoxGroup, and Flex Capital. The round brings Heron’s total capital raised to $23.3 million.
“Many industries still rely on human teams to process unstructured documents, creating a $300 billion global spend on business process outsourcing,” said Heron CEO Johannes Jaeckle. “Heron automates these workflows end-to-end.” According to Heron, In one case, an insurance company was able to automate over 80% of its inbound submissions while another lender cut its time from submission to decision by 60%.
The founders: Jaeckle, Jamie Parker (Chief Revenue Officer), and Dominic Kwok (Chief Technology Officer), came from high-growth fintech and tech firms including Taptap Send, Revolut, and Facebook.
Turning PDFs into Decisions
Heron’s product is an AI platform designed to integrate directly into the workflows of small and mid-sized financial services businesses, especially those without deep technical teams. Users forward inbound emails, upload PDFs, or integrate via API. Heron then classifies and extracts data, validates against internal policy, and pushes structured output into tools such as Salesforce, Quickbase, or Guidewire.
The platform relies on domain-specific prompts, internal testing frameworks, and feedback loops tailored to each customer and document type.
Deviating from intelligent document processing tools that stop at OCR or form recognition, Heron focuses on complete task ownership. If the system can’t automate a task, it flags the issue for human review. “Anyone who tells you they use AI to automate work with 100% accuracy is probably lying to you,” said Jaeckle. “We focus on clearly understanding where our software is successful and where humans still need to review”.
Vertical AI for Unstructured Data
Heron’s focus on smaller firms without engineering departments gives it a different growth profile from developer-first automation startups. Instead of offering modular APIs or open-ended tools, Heron delivers outcome-based automation designed to replace business process outsourcing in full. That makes the company’s competition less about other AI startups and more about incumbent service models: manual teams, external processors, or cobbled-together software stacks.
With the new funding, Heron plans to expand its presence in insurance, equipment finance, and SMB lending, while exploring adjacent sectors such as procurement and legal. The company will also scale its engineering and go-to-market teams, and invest further in internal tooling to increase customer capacity without increasing headcount.
While dozens of automation startups are building tooling for developers or back-office engineers, Heron is betting on end-to-end reliability as its differentiator. Its GTM strategy reflects that: Heron builds dedicated vertical teams to understand specific customer processes, fine-tunes models against representative documents, and delivers automation that can be deployed without technical setup.
Jaeckle describes their mission in operational terms. “We’re going workflow by workflow, industry by industry, replacing busywork with reliable AI so humans can focus on higher-value work involving judgment.”