When AI expanded from answering chat prompts to independently taking action across enterprise systems, software tools effectively became co-workers. Suddenly, security meant more than just governing a human employee’s identity.
For Juniper Square, which provides technology, data, and administration services for the highly regulated private funds industry, that change redefined what it means to keep data safe. In a sector where compliance and trust are paramount, moving fast with AI can't come at the expense of security control.
When the fintech company’s leadership championed the use of AI tools and agents across departments, the IT team faced an increasingly common question: How do you give teams the freedom to innovate with AI without giving away the keys to the kingdom?
For Juniper Square, the answer lay in a foundation it had already built.
How getting the identity basics right paved the way for AI
Long before AI tools entered the enterprise conversation, the IT team at Juniper Square was fighting a common operational battle: fragmented digital identities.
As the company quickly grew to more than 1,300 fully remote employees, its 12-person IT team wrestled with manual processes. Onboarding a new hire required wading through a 200-task Asana checklist just to provision app access. Offboarding presented a different headache. If a departing employee's account was simply deactivated, all of their shared client files would become inaccessible.
"Before Okta, it felt impossible for my team to stay ahead while addressing existing challenges," said Shen Ming Mah, Staff IT Systems Administrator at Juniper Square. “We knew we needed to standardize password management to keep accounts safe from phishing attempts without complicating the login experience."
With Okta, the team unified employee access across more than 175 apps under a passwordless, phishing-resistant framework, turned onboarding into an automated, role-based workflow, reduced manual offboarding risks, and streamlined compliance audits.
Without knowing it at the time, Juniper Square was putting into place the infrastructure it would need for a very different kind of user.
What happens when everyone wants AI
When Juniper Square introduced a full suite of AI tools across the organization, it triggered a surge of IT requests to connect AI agents and tools directly to internal databases, servers, and business software.
To maintain momentum without introducing security risks from shadow AI, the team established a formal process: Every new AI integration request goes through security review before approval. It’s the same scrutiny the team already applies to human access. This discipline ultimately revealed how to approach this new non-human workforce.
With a working approval model for AI tools in place, Juniper Square is now evaluating Okta for AI Agents to help it gain visibility and enforce strict access controls over what these autonomous tools can do.
"Being able to control what the AI agent account can or cannot do when it's interacting with other apps is really promising," Mah said. "Based on the foundation that I’ve built, I’m confident we can find a solution."
Bringing visibility to the non-human workforce
For Juniper Square, preparing for an agentic future is the natural continuation of the identity fabric they have already built. The lean IT team that secured access for 1,300 employees can now apply those lessons to AI. Because they automated their daily identity tasks, they have the time and infrastructure to focus on this transition. With a secure, automated foundation already running, Juniper Square is prepared for what comes next.
Want to hear more about how leaders are preparing for the agentic future? Join us at Oktane, September 22–24 in Las Vegas. Register now.
