Owl.co

What better way to spend summer than celebrating our #ScaleUPWeek2025 award winners? 🏆 If you missed out on this year's #ScaleUPAwards Gala, don't stress — over the coming weeks, we'll be sharing more about our winners and recognizing some of the boldest, most impactful and inspiring businesses scaling up across Western Canada. ⭐ First up is our Emerging ScaleUP of the Year award winner: Owl.co ⭐ Owl.co is a forward-thinking AI company committed to helping claims leaders transform insurance. Their technology equips insurers with human-first intelligence for ethical, informed, and swift decisions. In this video, Co-Founder and Chief Product Officer, Sohrab Merat, discusses the goals of Owl.co's future and the importance of connecting with other emerging entrepreneurs at events like ScaleUP Week. Congratulations, Sohrab and team on this impressive award win! 👏 Stay tuned as we share more from this year's ScaleUP award winners... 🔗 Stay up to date on #ScaleUPWeek news, photos and announcements at scaleupweek.ca --- ScaleUP Week 2025 is powered by Mount Royal University in collaboration with Growth Catalyst, McLeod Law LLP, Motiv and IVY. We couldn't do it without our sponsors! MNP, Export Development Canada | Exportation et développement Canada, Calgary Economic Development, The Opportunity Calgary Investment Fund, National Bank of Canada, Miller Thomson LLP, Virtuous Circle Counselling , Copoint, Viewpoint Investment Partners, ZayZoon, VistaVu Solutions, The New Network and the Digital Journal. #ScaleUPWeek #ScaleUPAwards #PathwayAwards #Entrepreneurship #Innovation

In insurance, compliance and ethics aren't merely buzzwords—they actually mean something. What are the green flags to look for in an enterprise AI solution? Here are three hallmarks that should be on your list if you're an insurance executive or senior manager for claims, legal, or SIU teams: 🤝 Accountable AI - Explainability: AI should delineate its logic—with audit trails, evidence, and citations—when generating claims insights. - Governability: Empower adjusters & other users with control over AI, which means approving recommendations and eschewing full automation. - Trainability: "Human-in-the-loop feedback" means users rate the relevance & quality of claims insights so the AI can continuously improve. ⚡ Effective AI - Precision: Document processors with a custom-curated vision-language model ensure best-in-class accuracy to make claims decisions with confidence. - Reliability: Built-in change guardrails and failsafe systems instill trust so claims teams feel empowered. - Speed: Processing documents and generating claims insights quickly and at scale liberates users with more capacity for critical thinking to enable empathy. 💛 Ethical AI - Compliance: Avoid legal headaches and look for toolkits that adhere to relevant regulations (CCPA, GDPR, HIPAA, and SOC 2). - Fairness: AI that's transparent about its processes and avoids risky predictive models helps mitigate bias and inadvertent discrimination. - Stewardship: Human-centric AI that enables faster & fairer claims decisions empowers executives to build a responsible future for the industry. These are the three tenets of Claims Intelligence. Combining effective technology with accountability measures and ethical outcomes via data & insights is how insurance leaders can create a new tomorrow and gain the appreciation they deserve. In short, powerful AI + impactful knowledge = Claims Intelligence. And it's the difference between supporting policyholders & growing business operations on one hand and falling behind & losing customer trust on the other.

AI is transforming insurance claims. But not all AI is created equal. For decision-makers, understanding the how of AI is as important as the why. Technical complexity can obscure real impact. So what are the key types of AI shaping claims operations? How do they work and where do they fit (or don’t) in a modern insurance ecosystem? Let's break it down for a clear, strategic lens: 🪄 Predictive AI Predictive AI analyzes historical data to forecast future trends or outcomes. For example, AI will "predict" claims outcomes based on past decisions. But it's ineffective for insurance claims; predictions can overlook nuance and context, risking inaccuracy, discrimination, and a lack of transparency. 🤖 Agentic AI Agentic AI can sometimes be predictive, but what uniquely defines agentic AI is when an AI "agent" stands in for a human employee. But agentic AI can overstep when automating delicate decisions without oversight. And we all know the frustrating feeling of dealing with AI agents.... ✍ Generative AI Generative AI "generates" outputs based on input prompts (e.g., ChatGPT). Generative AI benefits claims processing by providing actionable insights. For instance, with our generative-AI tool OwlAssist, investigators can ask questions about specific claims for real-time info to make fast & accurate decisions. 🔎 Deterministic AI Deterministic AI is rules-based, focusing on evidence, not predictions. It functions on predefined logic to analyze insurance claims in & of themselves. Unlike predictive AI that relies on historical-data patterns, deterministic AI provides accurate insights based on the claimant's actual information. Deterministic AI powers our insights tool, OwlSignal, for reliability and transparency, avoiding risky algorithmic biases. 📄 Document-Processing AI Document-processing AI reads documents to extract data, freeing investigators to leverage their creativity and intuition for handling complex claims. OwlVision, our document-processing tool, uses a custom-curated vision-language model (VLM) for best-in-class accuracy rates by combining textual & visual capabilities to understand document nuances. This VLM overcomes the limitations of some popular AI technologies that primarily depend on optical-character recognition (OCR), which is great for reading text but insufficient for messy handwriting or poor-quality scans. OwlVision uses OCR—but only to the extent that it's reliably accurate. That's because our unique VLM swoops in, so to speak, for optimal precision. *** To learn more, we published a report, "How AI Is Transforming Claims Processing in Insurance," that further delineates these details and outlines how insurers use AI & why AI matters for claims processing. Check it out here: <https://bit.ly/4mcOZZQ>.

Happy Independence Day! Let's talk freedom (i.e., privacy and regulations). Not the best day for spreading news, but this post is topical. And important. We hope you’re spending this weekend with loved ones, on a beach, or at a picnic. And if you don’t read this post right away (or at all), that’s okay! But in the spirit of independence, we wanted to outline our thoughts on issues that will impact the autonomy and sense of freedom that belong to U.S. citizens. 📜 The Near-Ban on AI Regulations—and Why It Matters 📜 Earlier this week, the U.S. Senate voted to strip away a provision in the House’s version of the One Big Beautiful Bill that would have banned states from regulating AI for ten years. And yesterday, a day before Independence Day, the House narrowly approved the Senate’s amendments. Why did the House originally want to ban, albeit not permanently, state-level AI regulations? The House felt pressured to address the issue of state-level regulations resulting in a patchwork of laws that could complicate compliance and thus, supposedly, hamper innovation. And it's true that the federal government has dragged their feet on AI regulations that would avoid such a potential patchwork. But without any AI regulations, Americans become susceptible to technology that could cause harm—especially in insurance, where stakes are high. People's livelihoods often hang in the balance. That’s why we value that insurance providers care about protecting their customers. After all, you’re in the business of helping people, and we want to support you on that mission. We support the great work that many states have already done to protect insurance policyholders and American citizens more broadly. Moving forward, we remain committed to providing AI for enterprise-level insurance providers that empowers corporate stewardship with Claims Intelligence—AI that’s accountable, effective, and ethical. Feel free to learn more about Claims Intelligence on our website <https://lnkd.in/g4vEDUsA>, including how we specifically prioritize best practices around explainability, fairness, and governance in our pursuit of accountability, effectiveness, and ethics.

In Case You Missed It | Garry Kasparov x Sean Merat "Human vs AI: The Future of Insurance Lies in Collaboration" At Insurtech Insights USA 2025, Garry Kasparov speaks on the opportunity that comes with AI, but the importance of the right questions being asked to improve the technology further. In conversation with Sean Merat, CEO & Co-Founder of owl.co, he laid out how true competitive advantage comes from collaboration - not confrontation with technology. 🎥 Watch the full session: https://lnkd.in/dXKt5x9d 🌏 Don’t miss more insights like with our weekly newsletter: https://lnkd.in/ddymNCY3 #InsurtechInsights #IIUSA2025 #GarryKasparov #Owlco #AIinInsurance #HumanMachineCollaboration #InCaseYouMissedIt Thomas Robinson

What good is document-processing AI for insurance claims? AI can liberate claims teams from manually handling documents, enabling human creativity & intuition to make complex decisions. But AI that merely processes documents only goes so far. What about *understanding* documents? How AI presents & categorizes info and generates claims insights makes all the difference in supporting genuine claimants and improving your bottom line. Okay, great, but how does document-understanding work? Claims teams rely on data-processing precision and accurate, comprehensive context in AI-generated insights. Implementing the wrong AI solution could yield huge problems. Let’s schematize document-processing AI and see how it all works. 🔠 Optical-character recognition (OCR): A technology that processes textual elements. It's great with plain text, but it has subpar accuracy with handwritten text and/or poorly scanned documents. Our document-understanding tool, OwlVision, contains some OCR—but only to the extent that it’s reliably accurate. 📖 Natural-language processing (NLP): Machine-learning technology that decodes and interprets text from natural languages, like words on a page. Unlike OCR, which processes individual letters to deduce words, NLP analyzes whole words and sentences to decipher context. 📸 Computer-vision processing (CVP): CVP is like NLP’s cousin, insofar as it does the same thing as NLP and in the same way, but it specializes in images instead of words. 💫 Vision-language model (VLM): An AI model that combines NLP and CVP for the best of both worlds—words and images—to fully process and interpret unstructured data of any type. OwlVision leverages a custom-made VLM that uses neural networks and user feedback to train its own knowledge on different types & presentations of data in claims documents unique to the insurance industry. 📄 Document-processing AI: The standard, high-level terminology for any technology that automates the processing of documents. Both OCR and a custom-made vision-language model are examples, though they’re not the same thing and don’t always co-exist. OwlVision, for instance, primarily relies on a bespoke VLM for a best-in-class document-processing rate of 97%. 📂 Document-classification AI: The "understanding" part of document-processing AI, how AI comprehends, interprets, and communicates what it processes. Document-classification AI segments and categorizes insurance documents for reliable claims knowledge. OwlVision deconstructs documents into their original components, labeling each document for streamlined retrieval; it constantly improves itself for reading & labeling documents and curating document categories based on insurance-specific context. 💡 Document-understanding AI: Augment document-processing with document-classification AI and you get…document-understanding AI, a superior technology for insurance companies. Now you know!

How can insurance carriers incorporate AI into their workflows? But, like, really, beyond the tactical steps and the operational instructions. The real question, of course, is: how will it impact claims decisions? Underwriting, fraud, and legal decisions? These decisions require complex and uniquely human critical-thinking skills. Intuition, creativity, empathy, and more. Your policyholders depend on it. When our own Sean Merat had the chance to interview chess legend Garry Kasparov, who famously played against IBM's AI in the 1990s, Kasparov was clear: the best decisions come from humans collaborating with instead of competing against or being replaced by AI. "The key is not human talent or the strongest machine. It’s a synergy, it’s a process, a superior process—the best collaboration wins. Decision-making is 100%, but you cannot reach 100% anywhere in the universe. It’s how to bring human contribution to the machine’s superior knowledge, which is probably 95–96%, and to make sure you push it as close as possible to 100%. This is the collaboration I believe in."

Excited to be a sponsor at this years ICA Conference in New Orleans! Looking forward to the sessions this year, come find us in September!

What an honor it was to share a stage with Garry Kasparov at Insurtech Insights in NY this week! Kasparov’s hallmark human + machine ethos in chess resonated with the insurance crowd and perfectly echoes owl.co's foundational belief: the real breakthrough isn’t humans competing against AI, but humans empowered by it.

We were honoured to accept the Emerging Scaleup of the Year Award from ScaleUP Week for the most-emerging scaling-stage tech company in Canada, with our own Shams Albayati on hand to receive the award at ScaleUP's gala event! Thank you ScaleUP for an incredible week and to all the other finalists for your incredible work and growth 🚀 We're excited for the next chapter in our collective journeys!