Read this online
Want to partner up with AI Insurer Brief, reply “Partner”
Welcome to AI Insurer Brief!
Hey, Fabio here!
In today’s Executive Series, I am joined by Hardeep Gulati, CEO of Duck Creek, following Duck Creek’s acquisition of Send.
Across insurance, underwriting teams are under pressure to make faster and more profitable decisions, yet many still rely on fragmented systems, duplicated rules, and information spread across multiple workflows. That raises a bigger question for the market:
What becomes possible when underwriting orchestration, core technology, and agentic AI are brought together?
Hardeep explained why the strategic value of the deal starts before AI, how tighter integration between underwriting and core could reduce duplicated implementation work, and how Duck Creek plans to use agentic and neurosymbolic AI across underwriting workflows.
Let’s dive in.
1. Hardeep, what does this deal now make possible for AI that was harder to do before?
The strategic rationale starts before AI.
Modern core systems are critical, but insurers also need underwriting to become faster, more efficient, and more profitable. Underwriters need the right information in one place so they can process submissions more quickly, assess the relevant risks, and make stronger decisions.
Send has focused on underwriting orchestration. It brings information together, automates routine tasks, and supports underwriters throughout the decision-making process.
The challenge has been deploying underwriting technology alongside core systems.
When an insurer has already implemented a core platform and then adds a separate underwriting solution, it may need to redefine the same products, rules, regulatory requirements, and information in both environments.
That duplication makes implementation expensive and time-consuming.
It has also been one of the major barriers preventing insurers from improving the underwriting experience.
By tightly integrating underwriting and core systems, we can reduce that duplicated work. Our objective is to enable insurers to deploy underwriting capabilities on top of Duck Creek in days or weeks rather than months or years.
AI takes that further.
Send was designed as an AI-native orchestration engine, using microservices and agents to coordinate underwriting workflows and information from multiple systems.
At Duck Creek, we have developed an agentic AI platform based on a neurosymbolic approach. We combine generative AI with business rules, knowledge graphs, APIs, and the data held within the core platform.
Our neurosymbolic approach adds the context, rules, and validation needed to make AI-supported processes more deterministic, auditable, and suitable for a regulated insurance environment.
The generative AI component can break down a workflow or decision, while the rules, APIs, and knowledge graphs within Duck Creek validate the output.
Together, we can create a comprehensive underwriting-to-core platform supported by an agentic AI layer with greater traceability, governance, and assurance.
Enjoying today's issue?
2. How do you see Send’s AI capabilities being integrated into Duck Creek in practice over the next 12 months?
We already have integration today.
Duck Creek and Send have common customers, and Send already integrates with multiple core systems. Existing customers can deploy Send through the integration currently in place.
The next step is to make that integration tighter through the agentic AI layer.
Rather than treating underwriting and core as two separate applications, we want to bring the workflows, information, and decision support from both platforms into one agentic experience.
We will apply Duck Creek’s neurosymbolic AI and AI-assurance approach across Send’s underwriting orchestration.
The goal is not simply to connect two systems.
We want to create a more unified underwriting-to-core experience in which data, rules, workflows, and AI-supported decisions work together.
3. Can AI also make the integration itself faster — for example around systems, data, and workflows — and if so, how?
Yes.
One of the benefits of our agentic AI platform is that it can change how we build the integration, not only what the end user experiences.
The traditional approach would involve connecting user interfaces, data, and individual system components through conventional integration methods.
Our approach is to bring information from Send and Duck Creek into an agentic layer.
Within that layer, the workflow, information, and elements of the user experience can be generated and orchestrated through AI.
For example, when an underwriter logs in, the AI could present the pipeline, identify which submissions have already been processed, route cases for approval, gather the relevant risk information, and highlight where human judgment is required.
In the agentic-first experience we are building, the underwriter would not need to navigate through an application and then separately ask an AI tool for assistance.
The AI would proactively show what requires attention, what has already been completed, and where human input is needed.
It could say:
“This is your pipeline.”
“I have already processed these submissions.”
“I am routing these cases for approval.”
“These are the cases where I need your input.”
“I have gathered the relevant risk data for you.”
That changes the role of AI within the application.
Instead of adding a chatbot or summarisation tool beside an existing workflow, we can use AI to orchestrate the workflow itself.
By using this design pattern, AI becomes part of the integration between Send and Duck Creek. It brings together information and workflows from both platforms without requiring every part of the existing user interfaces and data flows to be integrated traditionally.
That can make our own development, innovation, and integration work faster.
4. Looking five years ahead, what is the bigger vision behind bringing Send into Duck Creek, and what role do you think AI will play in how insurance is built and run?
Insurance is a highly regulated industry.
That creates a different challenge from sectors where AI can be introduced without the same level of scrutiny over how decisions are made.
Insurers need to demonstrate that policy, underwriting, and claims decisions followed the appropriate rules and regulatory requirements.
Today, insurers are already using some generative AI for internal support, developer productivity, and operational efficiency.
Over the next two to three years, I expect AI to move further into mainstream insurance workflows. That includes policy agents, underwriting agents, and claims AI.
For that transition to happen, insurers will need deterministic processes, human oversight, and AI assurance.
Human-in-the-loop is a critical requirement because people still need to review and approve important decisions.
AI assurance is equally important. Insurers need governance, auditability, and full traceability across the decision process.
Our core platform already captures product knowledge, business rules, and regulatory information in structured metadata.
We can map that core knowledge and those rules into the agentic platform, giving the AI the context required to validate its outputs against Duck Creek’s APIs and business rules.
That creates a neurosymbolic layer around the generative AI.
The generative AI can interpret the request and support the workflow, while the symbolic elements validate the answer against the rules and information held within the core platform.
The bigger vision is to make the underwriting-to-core environment agentic.
AI would orchestrate workflows, gather information, automate routine activity, and support decision-making. Insurers would retain human oversight and receive an auditable trace showing how the decision was reached.
That combination of AI, governance, traceability, and human control will be essential as insurers move from internal experimentation into policy, underwriting, and claims workflows.
An external study we referenced estimates that AI could create almost $1 trillion in value for insurers over the next decade through internal efficiencies, better underwriting, and the ability to serve broader markets.
We also referenced an estimated $30 billion to $50 billion market for AI technology in insurance.
The largest opportunity lies in applying AI to core insurance workflows with the assurance, governance, and traceability the industry requires.
See you on Friday!
Fabio Caravita
Founder, AI Insurer Brief
[email protected]
🏆 PARTNER UP WITH AI INSURER BRIEF
Reach insurance execs and decision makers weekly.
72% Exec / C-level / Decision-makers.
Limited number of partner spots open each quarter.
→ Reply “Partner” for details and we’ll get in touch.
☘️ AI OPERATING MODEL ADVISORY
For Insurers, Brokers, MGA, consultancies and tech providers, exploring AI or with a live AI mandate but limited clarity on governance, processes, operating model design and execution.
→ Reply “Advisory” for details and the team will be in touch.