Snowflake Startup Challenge 2026: Meet the 10 Semifinalists

In a year defined by the shift from chatbots to sophisticated autonomous agents, the competition in the 2026 Snowflake Startup Challenge was particularly fierce. We saw a variety of innovative startups across the globe, each leveraging Snowflake to push the boundaries of what’s possible with AI and data. After a rigorous review process, we’ve narrowed it down to 10 semifinalists who are looking to redefine industries ranging from cybersecurity and healthcare to geospatial intelligence and global supply chains.

These startups aren’t just building tools — they’re building agentic operating systems, security knowledge layers and digital twins of entire organizations. They’re making satellite imagery as searchable as a text document, deploying digital employees to manage risks or providing SMBs with the semantic infrastructure they need for reliable AI. 

Our semifinalists are at the forefront of the next wave of innovation, and we’re excited to see what contributions all of them make to the future of AI and data, regardless of who ends up walking away the winner of this year’s Startup Challenge.

We also want to thank all of the companies that took the time to submit an entry and share the exciting work you’re all doing. We appreciate your enthusiasm and effort. 

Introducing the 2026 Snowflake Startup Challenge semifinalists

Airrived

Airrived’s Agentic OS is a foundational operating layer that aims to move enterprises beyond simple AI chatbots toward building, deploying and scaling autonomous AI agents without the need for AI experts. Companies can fine-tune their models, deploy deep-reasoning agents for complex problems such as investigating a security threat, and orchestrate at scale so multiple agents can work together across different tools and departments. The AetherClaw governance and safety layer enables human-in-the-loop oversight to ensure autonomous agents follow company rules and security protocols. Snowflake is at the core of Airrived’s architecture, which leverages Snowpark, Snowflake Cortex AI and Secure Data Sharing as part of the scalable data and AI foundation of Agentic OS.

Auguria

Founded in 2022 by cybersecurity veterans, Auguria provides a Security Knowledge Layer to help streamline overtaxed security operations centers. The platform acts as an AI-powered data engine that sits between raw telemetry sources and analysis tools, using vector-based intelligence to automatically filter out up to 99% of security noise, according to Auguria. By transforming massive streams of logs into prioritized, explainable insights, Auguria helps organizations reduce SIEM costs, eliminate alert fatigue and provide a data foundation optimized for both human analysts and AI-driven security assistants.

Elait Health

Elait Health’s primary mission is to unify fragmented healthcare data — clinical records, claims, operational data and more — into a single, governed, AI-ready foundation for healthcare organizations. To tackle this challenge, its AI-powered health data management platform uses FHIR-native interoperability, which is designed to handle the full lifecycle of healthcare data. Elait significantly expanded its capabilities by merging with Airmeez, an AI engagement company, and creating a unified platform that enables healthcare organizations to deploy automated AI agents that perform patient scheduling, inquiries and follow-ups. Elait Health uses Snowflake as the primary data platform for its solution, taking advantage of multiple capabilities across ingestion, transformation, governance and AI.

Equity Data Science

Equity Data Science (EDS) provides a cloud-based investment process management platform designed to digitize and centralize the entire investment lifecycle for hedge funds and asset managers. By integrating qualitative research management with quantitative portfolio analytics and risk assessment, EDS acts as a single source of truth that replaces fragmented spreadsheets and manual workflows. Through its strategic partnership with Northern Trust’s Whole Office suite, the platform leverages AI and data science to help fundamental investors organize their proprietary processes, improve analyst productivity and transform intuition into a disciplined, measurable and scalable investment strategy.

Flexify

Flexify is AI for supply chain execution. It provides an agentic platform for supply chain and ops teams to help businesses move from simply identifying risks to actively resolving them through automation and intelligent reasoning. Built around the Snowflake Native App Framework, Snowflake Data Clean Rooms, Snowflake Marketplace and Snowflake Cortex AI, Flexify connects siloed systems such as engineering, sourcing and trade and uses autonomous agents to handle complex tasks. The solution can assess and score bills of materials, while its TariffPro module for trade compliance can automate HTS/ECCN classification. It also helps manufacturers and distributors adapt quickly to global regulatory shifts and supply chain disruptions by automating the search for alternative suppliers. 

HelloTwin

HelloTwin is building a semantic operating system that gives small and medium-sized enterprises the sophisticated data infrastructure they need to move beyond chatbots and deploy AI agents that understand how their specific company works. The platform connects business applications in minutes, creating a digital twin of an organization — a shared semantic model that aligns definitions, metrics and relationships across all tools. This unified foundation allows both humans and agentic AI to analyze fragmented data, generate trustworthy insights and execute coordinated, real-time actions across the entire business. 

LGND AI

LGND AI is a geospatial intelligence company that’s working to make Earth observation data from satellites and aerial imagery as queryable as text is for ChatGPT. With LGND Studio, a no-code collaborative platform, analysts can create and validate geospatial data sets with an AI-assisted workflow instead of relying on the slow, manual process of satellite image labeling. LGND Discover, built on Cortex AI, connects satellite imagery with other data sources like weather or points of interest to enable users to ask complex questions about the Earth and get visual answers in seconds, all in natural language. Developers can also use the LGND API to integrate observation data into their own models and applications. 

OpenBB

OpenBB is a fintech company that offers a modular, AI-fueled platform designed to help analysts and developers integrate and visualize vast amounts of financial data. Its OpenBB Workspace allows users to create interactive dashboards for financial data, while OpenBB Copilot serves as an AI research assistant that can summarize earnings calls or answer complex investment queries in natural language. Organizations can deploy OpenBB’s workspace on-premises or in a private cloud and unify all of their data sources — from internal data to public data and licensed third-party data feeds — in one secure, scalable system.

Twine Security

Overworked IT and security departments can breathe a sigh of relief thanks to Twine Security’s AI digital employees. Built using Cortex AI, the autonomous AI agents act as specialized virtual staff members that can plan, investigate and execute security actions instead of notifying a human to come in and handle them. Twine’s first digital employee, Alex, is an identity and access management expert designed to manage user entitlements, handle user access reviews and proactively address identity-related risks so IT and security teams can focus instead on high-level strategy tasks that need human intelligence.

Tynapse

Moving from chatbots to autonomous AI agents delivers significant productivity benefits — but also increases the risk of those agents making errors or being manipulated. To help mitigate this AI adoption risk for highly regulated industries, Tynapse provides a trust layer for enterprise agents and LLMs that sits between an organization’s applications and the AI models and is designed to verify interactions in real time. It provides agent guardrails to help detect and block security threats and hallucinations, and can protect personally identifiable information by masking or blocking sensitive data across all input and output channels. The platform can also automatically issue trust attestation sets and audit-grade proof to comply with global laws such as the EU AI Act.

Next up: Making the pitch and meeting the judges

In Round 2, each of these semifinalists will create an investor pitch video and join an initial interview with the judging panel. Based on this information, the judges will select three finalists that will present to our esteemed judging panel — now including Jon Herrick, Chief Product Officer at the NYSE; Snowflake’s Co-Founder and Vice President of Product, Benoit Dageville; Snowflake’s CMO, Denise Persson; and Capital One Ventures’ Partner,  Jon Barad — during the Startup Challenge Finale at Dev Day in San Francisco on June 4. 

Want to be there when we crown the 2026 winner? Register for Snowflake Summit and Dev Day to see the finale and explore over 500 demos, expert sessions, Q&As and hands-on labs designed to show you the latest Snowflake developments, jump-start innovative ideas and set you up for success in the era of agents and enterprise intelligence.

Congratulations again to all of the semifinalists, and best of luck!