DoiT transforms Opterrix’s insurance analytics platform with generative AI
Meet Opterrix
Opterrix is a cloud-native geospatial analytics platform built specifically for the insurance industry. The company integrates proprietary weather data, natural hazard peril scores, and predictive machine learning to help insurers optimize underwriting, respond to disasters, and manage risk in real-time.
With a vision to modernize and accelerate the insurance value chain, particularly in personal and commercial lines, Opterrix delivers mission-critical tools that empower underwriters and analysts to make smarter, faster decisions. “Our core mission is about reducing latency in insight,” says Niels Jorgensen, VP of Engineering. “We’re giving insurers the tools to act faster with confidence – especially when extreme weather hits.”
The Challenge
Severe weather is becoming both more frequent and more intense, placing enormous pressure on insurers to assess and respond quickly. Hailstorms, in particular, pose a unique challenge – they are localized, destructive, and unpredictable. After a storm, insurers must rapidly assess the event, compare it to similar ones, and activate response plans under significant time pressure. For Opterrix, enabling that process meant building tools to surface the most relevant historical hailstorm events on-demand. But engineers were consuming up to 10 engineering hours per case manually digging through data to find those matches – a slow, error-prone process that couldn’t scale.
At the same time, Opterrix’s lean team – just two full-time data scientists – was already stretched thin. While the company recognized that GenAI could revolutionize its workflow, the team lacked the internal bandwidth, production architecture expertise, and an implementation playbook to move quickly.
Opterrix needed a partner that could deliver the technical muscle and understand the pressures of working at startup speed with enterprise-grade precision.
The Solution
Opterrix engaged DoiT as a strategic partner to bring generative AI into production, with DoiT Cloud Intelligence™ guiding architecture, real-time spend visibility, and governance guardrails.
A consultative partnership built for agility
From the outset, DoiT approached the engagement not as a vendor, but as a strategic extension of Opterrix’s team. The accelerator program – designed to bring projects to life in a structured but flexible way – provided the perfect framework. Early discovery sessions focused on understanding Opterrix’s architecture, constraints, and goals in detail.
“DoiT didn’t try to apply a one-size-fits-all model,” said Jorgensen. “They customized everything to our business and platform – from the first workshop to the final deliverable.”
With clear alignment on outcomes, the DoiT and Opterrix teams focused on a high-impact target: a proof of concept for a hailstorm matching module using GenAI and multimodal vector embeddings.
From manual matching to machine intelligence
Previously, Opterrix engineers had to manually compare data and imagery from past hail events – a slow, inconsistent process and highly dependent on individual expertise.
DoiT helped Opterrix develop a solution that utilized GenAI to analyze images of hailstones and storm metadata, identifying similar historical events based on geospatial proximity and storm patterns. Leveraging Vertex AI and multimodal embeddings, the system understands visual and contextual features, transforming storm imagery and sensor data into searchable, vectorized intelligence in seconds.
Scalable, cost-efficient cloud architecture
In parallel, DoiT optimized Opterrix’s Google Cloud environment for both performance and cost. This included restructuring workloads to eliminate waste, implementing usage-based guardrails, and introducing DoiT Cloud Intelligence to automate efficiency gains.
Opterrix had previously struggled with balancing performance and cost as it scaled. DoiT’s deep knowledge of cloud-native architecture ensured Opterrix’ infrastructure could run more efficiently, without sacrificing responsiveness or uptime.
Governance and enablement
Because Opterrix works in a highly regulated industry, AI adoption had to be transparent, explainable and compliant. DoiT brought in frameworks for AI governance, documentation standards and reproducibility, ensuring not only that the model worked but that it could be audited and maintained over time.
Equally important, DoiT prioritized knowledge transfer through training, documentation and working side-by-side with engineers, ensuring Opterrix’s internal team could understand, extend and eventually lead future AI projects.
The Results
The Opterrix-DoiT collaboration didn’t just deliver a new feature – it unlocked a new way of working.
From 10 hours to 10 seconds
The AI-powered hailstorm matching module replaced a labor-intensive, 10-hour process with an automated solution that delivers results in under 10 seconds, freeing significant engineering hours and accelerating insurer response. This leap in efficiency now allows Opterrix to respond faster to weather events, giving insurers near real-time intelligence to expedite claims processing, mobilize resources and reduce losses.
50% faster time to market
By offloading the architectural design, model implementation, and infrastructure optimization to DoiT, Opterrix compressed R&D cycles by ~50% and increased delivery confidence.Features that once took months to research and deploy were now live in weeks, allowing the company to stay ahead of client demands and competitive pressures.
Jorgensen noted that the collaboration freed up critical internal capacity. “Knowing that DoiT had this handled gave us breathing room. It allowed us to stay focused on core business needs while still making forward progress on innovation.”
Cloud cost optimization with built-in scalability
Thanks to DoiT’s restructuring of Opterrix’s cloud infrastructure, unit economics improved significantly while preserving performance and uptime. By minimizing unnecessary resource use and applying intelligent scaling, Opterrix could scale usage with customer demand, without scaling cost at the same rate.
These efficiency gains continued even after the POC wrapped, as the platform’s new architecture sustained optimization across workloads.
Empowered teams and future-ready tech
Beyond technical deliverables, the project elevated Opterrix’s internal capabilities. Engineers gained hands-on experience with generative AI, understood how to integrate it into production systems and learned how to maintain governance in a regulated environment.
The AI module is now a cornerstone capability for Opterrix and a launchpad for further innovation. But even more importantly, the team is now equipped to extend that innovation independently.
“This wasn’t just about shipping a model,” said Jorgensen. “It was about transferring the skills, mindset and infrastructure needed to embed AI across our platform.”
What's Next?
With the GenAI foundation in place, Opterrix is exploring additional use cases – from predictive catastrophe modeling to more advanced underwriting tools – with governance and observability baked in.
As use cases grow in complexity, Opterrix plans to continue leveraging DoiT to design scalable, well-architected solutions with predictable unit economics and rapidly validate ideas through additional POCs.