Ajelix, the agentic AI platform used by more than 350,000 users across 150+ countries, today announces the acquisition of its own GPU compute infrastructure and the general availability of the Ajelix Enterprise Platform — a complete operational layer for organizations that need to run AI at scale, with full control over their data and processes.
The announcements arrive as businesses globally face a common problem: AI tools that work in demos but fail in operations, and infrastructure they don’t control, can’t audit, and can’t guarantee won’t be interrupted.
Ajelix Invests in GPU Infrastructure
The company has invested in its own GPU infrastructure to host and serve frontier open-source AI models like GLM-5.2 and Kimi K3, which now match the performance of closed frontier models on the majority of business benchmarks.
This is especially important news for organizations in regulated industries. Rather than asking a vendor to promise that data stays within certain boundaries, customers can verify it. This means no routing through third-party infrastructure and no foreign jurisdiction.
“We kept hearing the same question in every enterprise conversation: where does our data go?” – said Agnese Jaunosane, co-founder of Ajelix. “The honest answer, with most AI vendors, is that nobody can tell you with certainty. We decided to solve that problem from the infrastructure level up by owning the compute ourselves.”
The Enterprise Platform
Built on top of this infrastructure, the Ajelix Enterprise Platform is designed for organizations moving AI from experimentation into operations. It covers four core capability areas:
- AI Agents: Chat with agents that know your business, read your documents, use your tools, and follow your rules. Agents handle complex, multi-step tasks autonomously, creating presentations, Excel files, and dashboards, and escalating to a human when the decision is high-stakes.
- Workflows: Connect agents, tools, and people on a visual canvas with no code required. Workflows trigger on a schedule, an event, or a webhook and run end to end — branching, looping, and running in parallel across any data source or platform integration.
- Knowledge & RAG: Every document, policy, and dataset an organization connects becomes material the AI can reason over accurately. The system reads PDFs, Word files, emails, and scanned documents, finding meaning, not just keywords, and links every claim back to the exact source passage. This makes AI a tool that legal, finance, and operations teams can rely on.
- Security & Governance: Enterprise AI fails when it can’t be governed. Ajelix includes role-based access, complete audit trails, guardrails, and policy enforcement on every message in and out. Teams can catch sensitive data before it leaks, block banned topics, passwords, and keywords, control spend across projects and departments, and connect any LLM API through a single governance layer.
“AI enters daily operations, where it has to run like any other critical system: know your business, connect to your systems, follow your rules, and produce work your teams can use,” Jaunosane continued. “At this point, any organization have tried AI. The ones coming to us aren’t evaluating anymore, they’re trying to make it work in production.”
Deployment: Where the Data Lives
The enterprise platform is available in three deployment configurations to match any organization’s security policy or infrastructure preference:
- Ajelix Cloud — EU-hosted infrastructure, operated by Ajelix, ready to connect and go live immediately
- Self-hosted — the platform or models run inside the customer’s own environment, on-premises, on the edge, or in a private cloud; nothing leaves their infrastructure
- Cloud providers — available on AWS, Azure, or Google Cloud for organizations already committed to a provider
From 350,000 Users to Enterprise Operations
Ajelix has spent 4 years building toward this moment. The company started as an Excel formula tool, grew to 350,000 users without external funding, and launched its agentic AI chat in February 2026. The enterprise platform represents the natural extension of that trajectory: the same team that solved AI usability for hundreds of thousands of individual users, now solving AI control and reliability for organizations.
The platform has been developed alongside real enterprise clients spanning across financial services, legal, insurance, public sector, healthcare, logistics, manufacturing, energy, marketing, and real estate.
“We’ve been building AI since 2022,” said Jaunosane. “We’ve talked to hundreds of clients. The ones that successfully adopt AI are the ones that treat it like any other critical system: with real governance, real accountability, and real visibility into what it’s doing and what it costs.”
Why This Matters Beyond the EU
While data sovereignty has become a specific regulatory priority in Europe – the EU AI Act, the Cloud and AI Development Act, and national sovereignty programs across Germany, France, Austria, and the Netherlands have all raised the stakes – the underlying problem is universal.
Any organization in any jurisdiction faces the same operational challenge: AI tools that generate outputs nobody can audit, costs nobody can predict, and infrastructure that can be interrupted or changed by decisions made elsewhere.
Ajelix’s answer—owned infrastructure, open-source models, and a governance layer that gives organizations real control— is relevant whether the concern is GDPR compliance in Frankfurt, data residency requirements in the public sector, or simply the operational discipline expected of any enterprise software system.
Getting Started
The Ajelix Enterprise Platform is available now. Organizations can book a demo at ajelix.com/enterprise or contact the team directly.
Ajelix offers a structured implementation approach: assess, build, govern, adopt, support, staying through the first months of real production usage to ensure the solution actually works.
About Ajelix: Ajelix is an agentic AI platform built for work. Founded in Latvia in 2022 and bootstrapped to 350,000+ users across 150+ countries, Ajelix builds the platform, operates the infrastructure, and delivers the implementation expertise to help organizations run AI as a reliable part of their operations. Ajelix is a member of the NVIDIA Inception Program, Latvian Investment and Development Agency Incubator Program, OVHcloud Startup Program, AWS Activate, and Google for Startups.
This investment was supported by EU funding, the Latvian Investment and Development Agency, and Jelgava Municipality.
