Scout InsurTech Rising with FurtherAI

Aman Gour is the Co-Founder and CEO of FurtherAI, a company building the first AI Workspace for insurance that helps insurance teams automate complex workflows using AI Agents. Aman was interviewed by Chris Luiz, Co-Founder and CEO at Scout InsurTech, to discuss FurtherAI’s rapid early growth and how AI-native workflows could reshape the future of insurance.

Aman, you started FurtherAI in early 2024 and raised a Seed, then a Series A six months later. What changed in that window, both in the market and inside your company?

“We raised our Seed last April and then the Series A about six months later, which is obviously faster than the typical 18-month gap. A lot of that came down to a shift in how people started thinking about insurance and AI.

Insurance has always been called a data business, but most of that data lives in emails and documents. Historically, it was just stored as artifacts. You were not building insights on top of it because the technology was not there. Once large language models matured, especially around the GPT 3.5 era, that changed. Suddenly, you could actually process that unstructured data reliably.

The conversation moved from whether AI would matter in insurance to how fast it would transform workflows. Insurance is very document-heavy, and when AI can process those workflows end-to-end, you start to see real impact on expense ratios and even loss ratios because decisions are assisted at scale.

On our side, we launched with a handful of early customers and within six months grew from three or four partners to roughly 20 enterprises starting to use FurtherAI. At that point, the question became how we maintain the same level of high-touch delivery as we scale. Raising the Series A gave us the resources to go from five partners to 20 and soon 100+ partners without compromising outcomes. We did not want to overpromise and underdeliver. We wanted to overdeliver consistently.”

Where did you first find traction, and what was the wedge that allowed carriers and brokers to trust you?

“When we started, we did not come from insurance backgrounds. I had spent about a decade in AI, but not insurance specifically. So we literally drove around the Bay Area with donuts, walking into agencies and learning workflows firsthand.

That led us into the MGA market. MGAs tend to move faster than carriers or brokers when it comes to adopting new technology. They also view technology as a competitive edge. We started attending events like Target Markets and WSIA and saw early adoption there.

That became our initial wedge. We could demonstrate real improvements in metrics like submission to quote and quote to bind. Once we had those case studies, it reduced the perceived risk for carriers. MGAs had already experimented and validated the value.

Interestingly, I would have assumed claims would adopt AI first, but for us, underwriting moved faster. A big reason is delegated authority. MGAs and program administrators control underwriting workflows and could move quickly.”

You’ve described FurtherAI not as a software vendor, but as a transformation partner with skin in the game. What does that actually mean?

“One big realization with AI is that maintaining AI systems is very different from maintaining traditional software. Models are evolving rapidly. A new model comes out, and you may need to re-evaluate your entire stack to ensure accuracy does not regress.

Because of that, we felt it was unrealistic to hand customers a platform and expect them to manage the AI layer themselves. They have their own businesses and functions to run. So we leaned into a more forward-deployed model, similar to what companies like Palantir have done.

We embed engineers who configure the Workspace and continuously ensure our AI Agents perform as our partners' workflows evolve. That way, partners benefit from improvements without worrying about the operational burden.

Some people call it services as software or forward-deployed engineering. The label matters less than the outcome. We are not just giving a tool. We are giving teams an AI Workspace where AI Agents execute their standard operating procedures, and we take responsibility for making sure it delivers results over time.”

Insurance has always claimed to be a data business, but most systems only capture the final state of data. What changes when AI can capture the full workflow and decision trail?

“I do not think anyone fully knows yet, but directionally, it is a big shift. Today, systems record the final decision, like whether a policy was bound or a submission was quoted or claim was settled. But they do not capture the full journey. With AI, you can capture every step of the decision trail.

That creates a much more complete view of operations. You can understand what happened from the moment a submission arrived, why something was delayed, or why it was never quoted. That makes the entire value chain more efficient.

You also unlock new kinds of insights. Historically, a lot of context lived in the email threads and unstructured documents. If you wanted insights, you had to ask analysts to stitch together partial data. With full decision traces, business leaders can ask questions directly and get answers quickly.

It is similar to the shift from paper files to digital systems. That move created huge value because data became searchable and persistent. Capturing full decision trails is like taking another major leap forward. You have emails, documents, conversations, and decisions all connected. Every industry will feel that shift, but insurance is especially well-positioned because of how workflow-heavy it is.”

As AI Agents begin to talk to each other and orchestration becomes more autonomous, what do you see just over the horizon for insurance?

“One near-term example is cross-sell. Today, there is not enough cross-sell across lines of business, largely because data is siloed. If AI agents across lines can communicate, you start seeing opportunities surface automatically.

For example, an AI underwriting workflow in property could flag opportunities in D&O based on shared data, within appropriate data-sharing boundaries. Financial data gathered for one line could inform others.

If you extend that further, AI Agents across underwriting, claims, and finance can interact and generate real-time insights. Think about fast feedback loops. We have seen moments where external trends, like viral theft patterns, impacted underwriting quickly. Insurers who reacted fast limited losses. Autonomous orchestration could make those feedback loops much tighter.

Longer term, this could become proactive. AI agents continuously monitoring signals, triaging relevance to your portfolio, and surfacing actions. That might include recommending underwriting changes, alerting insureds, or prompting risk mitigation steps. With the right human checkpoints, you could have much more adaptive and responsive insurance operations.”