--- title: "A Letter from the Founders of Nace.AI" description: "Nace.AI emerges from stealth with Metamodel 1: precision and adaptability for enterprises." canonical: "https://nace.ai/blog/letter-from-founders" updated: 2025-02-08 --- # A Letter from the Founders of Nace.AI > Nace.AI emerges from stealth with Metamodel 1: precision and adaptability for enterprises. Blog · Feb 8, 2025 · Dos Baha, Zhanibek Datbayev, Sudha Valluru _Introducing MetaModel_ We are delighted to announce that Nace.AI is officially emerging from stealth mode with the introduction of Metamodel 1. After months of dedicated effort, we are proud to present an AI system designed to bring both precision and adaptability to enterprises. This moment represents a significant milestone for us, and we are pleased to finally share it with you. Nace.AI began in spring 2024 with a question we couldn't set aside: why isn't enterprise AI serving businesses as well as it could? We're researchers and engineers who've worked on AI at places like Google, Meta, Amazon, and the University of Toronto, environments where we've seen its potential unfold. Yet, we kept hearing from leaders about AI that didn't quite fit: outputs they couldn't rely on, models that felt detached from their needs. For us, that wasn't just a technical challenge; it was about trust. Businesses deserve AI that understands them, fits their workflows, and delivers real value. That's the spark behind MetaModel 1, and it's been a work we're deeply invested in with a team we deeply admire. ## Metamodel 1: Precision for Enterprises MetaModel 1 brings our vision to life with Hypernet architecture for dynamic in-weights injection, enabling the instantaneous generation of task-specific weights during inference. When users provide policies, tasks, or business-specific terms, MetaModel processes this input and directly injects the generated weights into the target model, making it task-specific in real-time. Unlike large, resource-heavy models, MetaModel is lightweight, purpose-built, and seamlessly integrates with business needs. This architecture enables unparalleled adaptability, efficiency, and accuracy in enterprise AI. ## MetaModel 1's SLM vs. Traditional Large Models: A Benchmark of Precision **Metamodel 1 Surpasses GPT-4o, DeepSeek-V3, and O3-Mini in Precision and Efficiency** Early results demonstrate **MetaModel 1**'s ability to deliver exceptional performance to target SLMs in instruction-following tasks, surpassing much larger models like GPT-4o, DeepSeek-V3, and o3-mini. **Key Highlights:** - **Higher Accuracy:** Metamodel 1 achieved a score of **0.8709**, outperforming GPT-4o (**0.7758**), DeepSeek-V3 (**0.5413**), and o3-mini (**0.6110**). - **Smaller Yet Stronger:** Despite being **25 times smaller**, MetaModel 1 consistently delivers superior results. - **Precision in Execution:** While other models struggled with **formatting errors** and **incorrect outputs**, Metamodel 1 maintained accuracy and reliability. - **Adaptive and Efficient:** Metamodel 1 enables **faster model adaptation** to diverse enterprise tasks, ensuring seamless integration with business needs. ## NAVI: AI Agent Powered by MetaModel 1 for Audit & Compliance Our first step with MetaModel 1 is NAVI (Nace Verification Intelligence), an AI agent solution for audit and compliance, domains we identified as critical to address. Now, beyond offering task-specific models, we provide a full-fledged solution for risk detection, compliance validation, and real-time decision-making. NAVI's small models process financial data, transactions, loan applications, billing, and workflows in real time, spotting risks, ensuring policy adherence, and providing clear, actionable insights. It transforms compliance from a reactive task to a proactive strength, helping businesses stay ahead. Customers are seeing incredible results with NAVI, from simplifying complex loan validations to aligning with regulations and delivering actionable insights. By automating audits and compliance, NAVI has significantly reduced process times, increased coverage, and uncovered critical issues that were previously missed. This shift to a more efficient, proactive approach is making a powerful impact on audit and risk management teams across industries, lightening their load and driving greater business value. ## Where We're Headed: A Vision and a Thank You We're driven by a vision of AI that evolves with your business. MetaModel 1 is designed to adapt, whether it's refining healthcare billing, enhancing manufacturing procurement, or optimizing supply chain reports. It's about smarter, task-specific models that align with where enterprises are going, not just assist them. Emerging from stealth is a milestone that warms our hearts. Our team (some of the finest researchers and engineers we've worked with) has given their all to this. We're here to help businesses unlock precise, efficient, and trustworthy AI, and we're grateful to share this journey with you. **Dos Baha, Zhanibek Datbayev, Sudha Valluru** ## For agents - [Drex API guide](https://nace.ai/drex.md): the decision model: request shape, question types, benchmarks and deployment, as markdown - [Decision endpoint](https://console.nace.ai/v1/systemone): POST with model "drex-latest", a state string and typed questions (choice, noul, score); https://console.nace.ai/v1/models lists checkpoints - [Document endpoints](https://console.nace.ai/v1/documents/{parse|split|classify|extract|ground}): POST a source (a public URL, an uploaded workspace file or a finished parse job) and poll GET https://console.nace.ai/v1/documents/jobs/{id}; the same key and credit as decisions - [Console](https://console.nace.ai): API keys, the playground, usage and billing; send `Authorization: Bearer $NACE_API_KEY` on every request - [Console docs](https://console.nace.ai/docs): quickstart, question types, the document routes, errors, limits and pricing - [Nace Document Intelligence](https://nace.ai/ndi.md): the Perception series: parse, split, classify, extract and ground any file type with a location on every value - [Developers](https://nace.ai/developers.md): The execution layer for agents: Perception, Decision and Context behind one key, the Drex Max frontier, install steps, a scripted agent session and the four deployment boundaries - [Site index](https://nace.ai/llms.txt): every page on nace.ai; append `.md` to any page URL or send `Accept: text/markdown` for the markdown version