Andrew O'Donnell
Build AI Products with Verisk's AVP of Model Product Management
Studied at Notre Dame University
Works at Core Wrk
Available August 12 at 11:00 AM UTC
Questions? Start chatting with this coach before you get started.
Andrew's Offerings
Custom hourly · $250/hr
Get help with AI Agents, AI Automation, and .
Andrew also coaches for AI Services and Build with AI. View all.
Andrew’s AI for Product Development Qualifications
Experience level: Vice President
With a Ph.D. from Notre Dame and extensive experience in AI-native product development, I am passionate about helping clients harness the power of AI to drive business outcomes. As the founder of ClaimCode, I built a cloud-native, multi-tenant, SaaS application that processes PHI in a HIPAA compliant manner to give small, independently owned wellness practices the revenue cycle intelligence necessary to balance theri books and make sure their are fairly compensated. I have also developed find-your-fit.org as an online community for fitness and wellness enthuiasts and professionals to discovery and be discovered with a focus on each individual's unique fitness journey. As the Principal at Core Wrk, I focus on bridging the gap between AI tool access and real-world deployment, developing platforms like HelmFast and Employee Number Two to empower non-technical founders. My role as Assistant Vice President in the Product Management team at Verisk has equipped me with the skills to lead AI adoption initiatives, having successfully implemented AI systems that reclaimed significant resources and improved operational efficiencies. Whether you're looking to define, scope, deploy AI systems, or build your first AI-powered product, I'm here to guide you through the process. Let's connect and explore how we can achieve your AI product development goals together!
Andrew can help with:
AI Agents
AI Automation
AI Product Design
AI Roadmap Development
Feature Prioritization
LLM-Optimized Code Base
Product Strategy
Prototyping with AI
QA
Scoping & Requirements
User Research & Insights Automation
Andrew also coaches for AI Services and Build with AI. View all.
About Andrew
Andrew O'Donnell holds a PhD from the University of Notre Dame and has spent over a decade at Verisk Analytics working at the intersection of catastrophe modeling, data analytics, and enterprise technology. He founded ClaimCode, an AI-native healthcare SaaS company, and Core Wrk, LLC, an AI product consultancy, both built using AI-enabled tools and applications. He is a Verisk Certified Catastrophe Risk Modeler (VCCRM), a Lean Six Sigma Green Belt, and an AWS-certified cloud practitioner.
Why do I coach?
Andrew has been coaching and mentoring informally for most of his adult life: first on the Notre Dame cycling team, then at a triathlon gym in South Bend, Indiana, and more recently at Equinox. The sport has changed, but the work has always been the same: figure out where someone's effort and resources aren't aligned, then fix that. That habit of mind carried into his career. When he took on the Research Operations role at Verisk, a big part of what drew him to the work was the chance to build training and enablement programs for a 150-person research team. Brilliant scientists shouldn't have to spend their time on operational friction. He wanted to fix that. AI coaching is that same instinct in a new domain. Most people have far more capacity to deploy AI effectively than they're currently getting. Closing that gap to make AI feel accessible is the work.
Work Experience
Founder & CEO
ClaimCode
May 2025 - Present
I founded ClaimCode to bring AI-powered revenue cycle intelligence to small physical therapy and wellness practices, taking the product from concept to production in under 10 months without a traditional engineering team. I built and operate a fully autonomous outbound GTM engine (Apollo → HubSpot → AI classification → Gmail) with no manual intervention, as well as agentic content and inbound marketing agents. The platform is built on a HIPAA-compliant, multi-tenant architecture with privacy-first data modeling, AI claim coding, and real-time denial-risk scoring. ClaimCode is my live proof point for coaching clients on how to build and operate AI-native products responsibly and efficiently as a non-engineer founder.
Principal
Core Wrk
September 2024 - Present
Core Wrk is my vehicle for AI-native product development and consulting, focused on closing the gap between AI tool access and real-world deployment. I am developing HelmFast, a bring-your-own-cloud (BYOC) platform enabling non-technical founders and product teams to deploy locally-built agentic workflows to cloud infrastructure they already own. I am also building Employee Number Two, an open-source agentic startup advisor that produces tangible deliverables pitch frameworks, competitive analysis, positioning copy rather than generic guidance. Core Wrk reflects my core coaching thesis: the highest-leverage skill in AI today is knowing how to direct agentic systems toward business outcomes, not how to build them from scratch.
Assistant Vice President
Verisk
April 2023 - Present
Hiring Manager
AVP, Model Product Management — Verisk Analytics (AIR Worldwide) (2025–Present) I lead product strategy and prioritization for North American catastrophe risk models at one of the world's leading analytics firms. In this role I built and deployed the Model Insight Utility — an agentic AI pipeline that replaced a manual Salesforce review workflow and reclaimed 32 person-hours per week across research, support, and marketing teams. This role is the direct source of my coaching perspective: I teach clients how to define, scope, and deploy AI systems as a product discipline. AVP, Research Operations — Verisk Analytics (AIR Worldwide) (2021–2025) I founded and ran a new operations function for a 150-person research organization, leading its technology transformation from the ground up. Key outcomes included a $1M+ annual cost reduction from a 2.5 PB AWS cloud migration, 50–80% runtime improvements on HPC workloads, and a 75% reduction in audit cycle times through process redesign and automation. I also led a formal enterprise knowledge management RFP and deployed AI-powered enterprise search across the organization. This background informs how I coach leaders on structuring AI adoption initiatives inside complex, regulated organizations.
Cycling Instructor
Equinox+
July 2018 - March 2020
Leading indoor cycling classes focused on an authentic road riding experience with tried and true training principles. Oh yeah, and the music is pretty fun too. Presentations, Sports Coaching and +3 skills
Earthquake Engineer and Model Developer
Verisk
December 2014 - April 2021
Hiring Manager
Developed high-provide earthquake risk models for all earthquake-prone geographies including United States, South America, Southeast Asia, New Zealand, and Australia.
Education
Notre Dame University
Doctor of Philosophy (Ph.D.), Civil Engineering
Activities and societies: EERI, ASCE A macro-element modeling approach for unreinforced masonry structures (URMs) that is compatible with the OpenSEES modeling framework is developed and validated against and extensive suite of experimentally tested URM structures. The modeling approach is applied to a probabilistic fragility assessment of different URM building typologies. A catalog of prototypical URM structures characteristic of the buildings that exists in the Northeastern United States is compiled and a series of dynamic time history analyses are performed to determine probabilistic fragility curves and damage functions that accurately describe URM behavior. Ultimately, these fragility curves and damage functions can be used for seismic risk and portfolio loss assessment of these types of building structures. Athletics, Sports and +2 skills
Notre Dame University
M.S.C.E., Civil Engineering
Activities and societies: EERI, ASCE An experimental evaluation of ground motion scaling methods for the nonlinear analysis of structural systems was performed. A specimen was designed to be nonlinear and reusable through the use of a novel connection that was designed and fabricated at the University of Notre Dame. The repeat-ability of the specimen was shown though a myriad of validation experiments. Selected ground motion scaling methods were applied to an extensive suite of ground motion records and conclusions were made for a preferred ground motion scaling approach. Athletics, Sports and +2 skills
2 Reviews
Overall Rating