Snowflake Startup Spotlight: Ekai

Ekai’s real technical bet is not just model generation, but context production. For enterprises, agentic AI often fails less because of the LLM and more because the surrounding data architecture is fragmented: business definitions, lineage, entity relationships and validation rules live in separate tools, spreadsheets or people’s heads. A system that can infer and maintain that context becomes a critical dependency for any downstream natural language or code-generation workflow.

This also changes where the work sits in the stack. By running inside the customer’s Snowflake environment, the deployment model reduces the usual data-movement and security friction that can stall analytics and AI projects. For architects, that means the product is not simply an external SaaS layer; it is an in-warehouse extension that must align with existing warehouse governance, access controls and compliance boundaries. The integration choice is therefore as important as the AI capability itself.

The operational trade-off is governance versus speed. Automating catalogs, glossaries, lineage maps and ERDs can compress modeling cycles, but enterprises still need clear ownership over what the generated artifacts mean and when they are production-ready. In practice, the value is highest when business users can prototype quickly and IT can validate, harden and operationalize the result without redoing the discovery work.

For teams building AI-ready data products, the broader lesson is architectural: agentic systems need a continuously maintained semantic layer, not just cleaner prompts. The winning pattern is likely to combine automation for discovery and documentation with human control over standards, approvals and release management.

Welcome to Snowflake’s Startup Spotlight, where we ask startup founders about the problems they’re solving, the apps they’re building and the lessons they’ve learned during their startup journey. In this edition, meet Ekai Co-Founders Mo Aidrus, CEO, and Hussnain Ahmed, Chief AI Officer, who are applying their 20+ years of experience in data engineering and agentic AI, context and ML Ops, and enterprise cloud infrastructure solution engineering to build a more efficient infrastructure for enterprises’ agentic data needs.

What inspires you as founders?

We are obsessed with using data and AI to make the enterprise world more efficient. We’re now in the third year of the generative AI adoption wave, and there’s a pattern emerging: Companies use applications like ChatGPT, Claude AI, Grok and so on, or they build custom applications using LLM APIs from OpenAI, Anthropic and Google. They get decent initial results on generic tasks, then hit a wall when trying to use these tools for anything that requires understanding their actual business. The problem isn’t the models. It’s context. Or more precisely, it’s the complete absence of structured, queryable, machine-readable context about how an enterprise actually works. Agentic data systems are worthless without enterprise context.

How does Ekai aim to solve this problem?

Despite all the available tooling, the process of understanding physical data, contextualizing it and transforming it into business-ready insights remains painfully slow and manual. That’s why most organizations barely scratch the surface of what their data platforms can do, and why expensive data catalogs often go underutilized. Our platform seeks to change this by offering a “business data lab” for business analysts and analytics engineers to build various data models without extensive IT involvement. Ekai connects to physical data, learns patterns, creates logical data models and generates code that merges physical data with business concepts. The goal is to empower business users to prototype and test models before engaging IT for production deployment. Ekai’s AI chat agent assists business users in defining semantic models by asking intelligent questions based on its understanding of the physical data. The goal is to empower business users to prototype and test models before engaging IT for production deployment. This process helps narrow down business requirements, similar to how data architects conduct workshops to build a business requirement document (BRD). Our conversational approach makes data modeling accessible to business users and, most importantly, documents tribal knowledge.

What makes your approach to democratizing data models and insights different from other solutions in the market?

Our focus is to accelerate data innovation and time to value. Ekai’s proprietary technology can:

Automatically infer, catalog and build entity relationships scalably
Catalog business definitions and processes that are reflected in existing SQL code
Build accurate data product prototypes for business users
Enable accurate and reliable data discovery in natural language
Continuously build and update documentation for physical, logical and semantic data
Build specifications for downstream AI applications
Make a company’s dark data agentic AI ready

We connect to your data warehouse, automatically generate enterprise mind maps and ontology/entity-relationship diagrams (ERDs) to understand relationships, capture business context through our semantic modeling and produce all the artifacts that downstream AI applications need: data catalogs, business glossaries, metrics definitions, lineage maps and validation rules. It’s the entire context layer, generated and maintained automatically. With Ekai, what used to take three to six months to spin out data products can now be accomplished in three to four hours with AI. Business users don’t feel trapped, and they can create semantic models with ease with the appropriate enterprise context.

How has the Snowflake Native App Framework influenced your company’s growth and development strategy?

The Snowflake Native App Framework is pivotal for us, because it helps us solve two critical adoption barriers. First, by executing entirely within the customer’s Snowflake account, all data and metadata remain in their security perimeter, eliminating extraction concerns and accelerating compliance approvals, especially for regulated industries. Second, it positions Ekai as a native extension of Snowflake rather than another external AI tool. That means customers can deploy automated data modeling directly within their existing warehouse infrastructure with minimal friction. This transforms our go-to-market strategy from a complex integration sale into a streamlined activation experience that drives faster adoption and product-led growth. We are very excited to work with Snowflake so we can reduce procurement overhead and obtain billing and metering right out of the box. Co-selling via the Crossbeam platform has already started.

What’s on the horizon for you and Ekai?

As AI begins to lead analytics, this level of automation won’t just be a competitive edge — it will be a necessity. AI systems can only deliver real value if they’re fed with well-contextualized, business-ready data. That’s what we deliver: content that generates accurate, consistent and complete output. With Ekai, enterprises get accurate and reliable agentic code generation and natural language querying, while automatically cataloging and documenting their code. Learn more about Ekai’s solutions at http://www.ekai.ai. Ekai is currently available in Snowflake’s private listing. Please reach out at https://ekai.ai/#contact to get started. If you’re a startup building on Snowflake, check out the Powered by Snowflake Startup Program for info on how Snowflake can support your goals, and be sure to enter the 2026 Snowflake Startup Challenge!

Original Post

Enjoyed this article? Sign up for our newsletter to receive regular insights and stay connected.