Sign in
Welcome! Log into your account
Forgot your password? Get helpPrivacy PolicyPassword recovery
Recover your password
A password will be e-mailed to you.
GEN – Genetic Engineering and Biotechnology NewsHomeTopicsArtificial IntelligenceAI “Council of Models” Improves Workflows and Outcomes
Credit: Andriy Onufriyenko/ Getty Images
To maximize the benefits of AI, biopharmaceutical manufacturers need to take an end-to-end systems engineering approach to their data, preparing it for AI while orchestrating the right models for each stage of a workflow. Relying solely on a single foundation model is often insufficient, particularly as complexity increases and therapeutics advance from pilot stages into production.
No single model excels at every task. Running the same engineering process through different AI models often produces significantly different results. Even repeated runs on the same model can yield inconsistent outputs, Farshid Sabet, CEO and co-founder of Corvic AI, tellsGEN.
Variability may be acceptable for low-risk activities, but it becomes problematic when engineering diagrams, flow directions, operational relationships, and other complex data are involved. “Small inaccuracies can compound quickly,” he cautions, leading to unreliable results in production environments.
Corvic AI recently benchmarked leading frontier AI models against Corvic V5’s workflow orchestration platform, assessing their ability to extract piping and instrumentation diagrams (P&IDs) into XML files.
“For general text generation, today’s frontier AI models perform remarkably well and the differences between them are relatively small,” Sabet says. “But engineering workflows introduce an entirely different level of complexity.”
The benchmark found that relying on foundation models alone often resulted in inconsistencies, hallucinations, and poor repeatability. Corvic addresses these challenges by combining semantic data preparation with workflow orchestration that coordinates multiple AI models, validation steps, retrieval, and enterprise context to improve reliability.
Rather than replacing frontier models, Corvic’s platform integrates and orchestrates them, selecting the best model for each stage of a workflow based on the task, performance requirements, and cost.
Like wild horses
As Sabet says, “AI models are like wild horses. They’re incredibly powerful, but they need guidance, structure, and context before they can consistently solve complex enterprise problems.”
Typically, AI developers focus on improving the models themselves through training, fine-tuning, or prompt engineering while assuming enterprise data is already AI-ready. Corvic, instead, focuses on organizing enterprise knowledge through a semantic layer that enables AI systems to understand relationships across engineering documents, databases, diagrams, and operational systems.
“We work with the data independently of whether it’s manufacturing, chemistry, or biology,” Sabet says. “The data has to be organized in a way that allows AI models to recognize context and relationships. That’s the semantic layer.”
Once enterprise knowledge is structured appropriately, organizations can intelligently orchestrate multiple AI models throughout a workflow rather than relying on a single model for every task. Sabet refers to this approach as a “council of models,” where each model contributes its strengths to improve overall accuracy, repeatability, and efficiency.
A former Intel executive, Sabet founded Corvic AI to help organizations operationalize AI across complex enterprise environments. Today, the company works with manufacturers, life sciences organizations, and other enterprises to transform fragmented operational knowledge into reliable AI workflows that improve productivity and decision-making.
For organizations evaluating AI platforms, Sabet recommends looking beyond benchmark scores and considering three factors: how enterprise data is secured and governed, whether the platform intelligently matches AI models to different stages of a workflow, and how success will be measured through meaningful productivity outcomes.
To maximize the benefits of AI, Sabet reiterates, biopharmaceutical manufacturers “need to look at their data from a systems engineering perspective.
“The future of enterprise AI isn’t about finding one perfect model,” he says. “It’s about intelligently orchestrating enterprise data, semantic understanding, and specialized AI models into repeatable workflows that organizations can trust.”
InsightsArtificial intelligenceBiopharmaceuticalsCorvic AICorvic V5Also of Interest
CloudScope Enables Continuous Remote Monitoring of Brain Activity in Freely Moving MiceSmarter Cell Culture Starts with Better MediaRevvity Signs Agreement to Acquire Human Cell DesignTop 10 Contract Development and Manufacturing Organizations 202610 CDMO Up & Comers 2026Harnessing Real-Time DataRelated Media
Data Integrity as the Foundation for AISmall Molecules, Big Expectations: How CDMOs Are Helping Sponsors Navigate Complexity, Speed, Scale-Up, and SustainabilityBeyond the Technology: AI Readiness in Regulated LaboratoriesAACR 2026 Video Update: Cancer Research Edges Toward an AI-Driven EraReporting Live from JPM 2026: Alex Philippidis and Jonathan Grinstein, PhDAI in Protein Design: Hype vs. Reality Explained by David BakerTop 5ResourcesRecommended For YouPodcast
Touching Base
Touching Base is the dynamic podcast series from the editors ofGEN. Each episode features a rotating case of senior editors—including John Sterling, Kevin Davies, Julianna LeMieux, Alex Phillippidis, Uduak Thomas, Corinna Singleman, and Fay Lin—who delve into emerging stories, exchange ideas, and debate the latest trends in biotech. Additionally, they talk to some of the leading voices in the industry about what's now and next.Start listening today!
Stay up to date with the lasted episodes of Touching Base bysubscribing to theGENPodcast Newsletter
Copyright © 2026 Sage Publications or its affiliates, licensors, or contributors. All rights reserved, including those for text and data mining and training of large language models, artificial intelligence technologies, or similar technologies.
MORE STORIES
