# Chris Lovejoy on Building Domain-Native AI Organizations _Chris Lovejoy of Notius Labs discusses the critical role of domain experts in AI product development, outlining three key organizational models: Oracle, Evaluator, and Architect._ **Updated:** 2026-08-22 **Published:** 2026-05-16 **Source:** https://www.startuphub.ai/ai-news/artificial-intelligence/2026/chris-lovejoy-on-building-domain-native-ai-organizations --- Chris Lovejoy, Founder of Notius Labs, shared insights on building effective AI organizations at a recent AI Engineer Europe event. His presentation, "How to Leverage Domain Expertise to Build Better AI Products," emphasized the critical role of domain experts in creating successful AI solutions. Lovejoy argued that the organizational structure and the integration of domain knowledge are more crucial than the underlying AI models themselves. AI Integration ChallengesDriverorganizations struggle to effectively integrate AI into productsFrom the article 3 mentionsLovejoy argued that the organizational structure and the integration of domain knowledge are more crucial than the underlying AI models themselves.requiresDomain Expertise CrucialCoredomain experts are more vital than AI models themselvesFrom the article 6 mentionsHis presentation, "How to Leverage Domain Expertise to Build Better AI Products," emphasized the critical role of domain experts in creating successful AI solutions.enablesDomain-Native OrgContextorganizing teams around specific business knowledgeFrom the articleUltimately, Lovejoy concluded that building a successful domain-native AI organization requires a strategic approach to integrating and empowering domain experts throughout the product lifecycle.includesOracle RoleCoreembedding domain expertise directly into AI applicationsFrom the article 9+ mentionsLovejoy outlined a framework for organizing teams around domain expertise, proposing three key roles: Oracle, Evaluator, and Architect.Evaluator RoleCoreFrom the article 9+ mentionsThe Evaluator role focuses on defining and measuring AI quality, establishing metrics and systems to track performance.Architect RoleCoreFrom the article 9+ mentionsFinally, the Architect role designs systems that enable continuous learning and improvement through automated feedback loops.Better AI ProductsEffectFrom the article 9+ mentionsHis presentation, "How to Leverage Domain Expertise to Build Better AI Products," emphasized the critical role of domain experts in creating successful AI solutions.shown byCase StudiesOutcomeGranola, Tandem, and Anterior demonstrate successFrom the article 3 mentionsLovejoy illustrated his points with three case studies: ## The Domain-Native AI Organization Lovejoy outlined a framework for organizing teams around domain expertise, proposing three key roles: Oracle, Evaluator, and Architect. The **Oracle** role involves directly embedding domain expertise into the AI application, often through prompt engineering or curated data. The **Evaluator** role focuses on defining and measuring AI quality, establishing metrics and systems to track performance. Finally, the **Architect** role designs systems that enable continuous learning and improvement through automated feedback loops. He stressed that the choice of approach depends heavily on the specific use case and scale of the AI product. For instance, a startup might begin with a single domain expert acting as an Oracle, but as the product scales and faces challenges like the need for more nuanced quality assessment or the handling of diverse customer needs, the organization may need to transition towards an Evaluator or Architect model. ## Addressing Common Mistakes in AI Integration Lovejoy highlighted common pitfalls companies encounter when integrating AI, including not hiring domain experts early enough, selecting the wrong type of expert, or failing to integrate them effectively into the organizational structure. He asserted that the core challenge lies not just in having the right models, but in operationalizing expert judgment around them. He presented a decision tree to guide organizations: if AI quality can be measured objectively, the focus shifts to defining and measuring performance. If manual iteration is fast enough, a single domain expert as an Oracle might suffice. However, if manual iteration is too slow, the organization should aim for an Architect role that can automate improvements. ## Case Studies: Granola, Tandem, and Anterior Lovejoy illustrated his points with three case studies: - **Granola:** This company, which uses AI to generate meeting notes, benefited from Jo Barrow, a writer/journalist with extensive research skills. She acted as the primary gatekeeper of AI quality, with her role as an Oracle working well due to the subjective nature of meeting notes and the single core output. - **Tandem:** This medical AI scribe company hired a medical doctor, Roy Pekny, as an early employee. As the product scaled and needed to support a long tail of prompt customizations across various specialties and geographies, the company transitioned to a decentralized Oracle model, hiring doctors from different regions and specialties to manage these nuances. - **Anterior:** Lovejoy's personal experience at Anterior, a prior authorization AI startup, demonstrated a progression from an initial Oracle role (building prompts and code) to an Evaluator role (defining metrics, building review dashboards, and hiring clinicians). Due to the complexities and variations in prior authorization rules, they eventually needed to move towards an Architect role to automate improvements. These case studies illustrated how the initial approach might evolve as the product and organization scale, emphasizing the need for adaptable roles and a deep understanding of how to integrate domain expertise effectively. ## Key Takeaways for Organizations Lovejoy summarized his advice into three key principles for leveraging domain experts: 1. **Define a Principal Domain Expert:** Assigning clear accountability to a single individual can accelerate decision-making and product development by avoiding consensus-building by committee. 2. **Give Them Ownership:** Domain experts should be deeply integrated into the product development process, not merely consulted. True ownership allows them to build differentiated products. 3. **Hire for Breadth and Evolve Roles:** Seek individuals with relevant domain expertise and a breadth of adjacent skills. Start them in an Oracle capacity and allow their roles to evolve into Evaluator or Architect as the product and organization grow. Ultimately, Lovejoy concluded that building a successful domain-native AI organization requires a strategic approach to integrating and empowering domain experts throughout the product lifecycle. --- Original analysis from [startuphub.ai](https://www.startuphub.ai), the #1 AI startup directory. © StartupHub.ai. All rights reserved. You may not republish this article in full without a license. Search engines and AI research tools may crawl and summarize for reference. Bulk reproduction or model training on this content requires a license. See https://www.startuphub.ai/terms.