Client story Giles AI Builds a Secure Cloud Foundation in 11 Weeks

Advancing artificial intelligence for medical research

Giles AI stands at the forefront of a major shift in how medical research happens. The company uses advanced artificial intelligence to help researchers find answers faster, with the ultimate mission of saving time to save lives. However, the company faced a tight deadline.

They targeted an Oct 2025 release for their public product and knew their existing setup couldn’t support their ambitious growth plans. To meet this challenge, they needed a solution provider who could move fast and build a foundation ready for the future. Giles AI chose Insight to help them migrate to Google Cloud.

The challenge

Giles AI faced urgent time pressure. With a public product launch targeted for Oct 2025, the team realized their current technology mix would limit their growth. They had started with a generic infrastructure that worked for early testing but lacked the power needed for large-scale operations. Specific challenges included:

• Need for specialized AI hardware: Giles AI required a platform and infrastructure that were optimized for AI workflows.
• Strict compliance: Operating in the healthcare sector meant they had to meet strict standards like SOC 2 and GDPR.
• Operational risk: The engineering team needed to eliminate configuration drift among their QA, Dev, Demo, and Production environments to ensure reliability. They needed to make sure the environments were all set up the same way all the time.
• Cost predictability: To scale rapidly and cost effectively, they needed a clear view of their potential cloud costs.

“Our cloud foundation was mission‑critical for our launch,” says Kiran Dhaliwal, COO at Giles AI. “It set Giles AI up to grow aggressively without costly re‑architecting.”

The solution

To solve their business challenges, Giles AI decided to work with Insight to build a secure, scalable Google Cloud environment. By choosing Google Cloud, Giles AI was able to take advantage of Google Cloud’s Tensor Processing Units (TPUs) for their complex AI workloads and the best-in-class Vertex AI tools, which matched their needs for rapid iteration and agile workflows.

“We knew Google Cloud’s TPUs and Vertex AI platform would give us the flexibility for experimentation and the scalability that we needed,” says Dhaliwal. “Vertex AI would let us experiment in a way we couldn’t before. Because the market moves really fast, we have to stay on top; we have to stay in front of that movement, and there’s no better way to do that than through collaboration with Google Cloud and Insight.”

As part of that collaboration on Vertex AI, Insight delivered a standardized landing zone based on proven best practices, allowing Giles AI to have a secure setup from day one.

Using a phased approach

To manage complexity, Insight split the project into two phases:

• Foundation: First, they locked down the basics of networking, identity, and security guardrails.
• Minimum Viable Cloud: Once the base was in place, they deployed the application layer. This let the team test real workloads and validate security controls before going into production.

Managing infrastructure with Terraform

Insight helped Giles AI adopt Terraform to manage their infrastructure. This moved the team away from manual clicks and toward a system where every change is version-controlled with Infrastructure as Code (IaC). This shift ensured that the QA, Dev, Demo, and Production environments all behave exactly the same way.

“Because there’s only one source of truth, maintained by our Terraform IaC tool, every change that we make to our cloud environment is highly auditable,” says Andrei Niculescu, Infrastructure Lead at Giles AI. “Everything is version-controlled and deployed consistently across all environments, which is essential to ensuring that what arrives in Production has been thoroughly tested and adheres to company and industry standards.”

Designing security into the network architecture

Insight designed security directly into the network architecture. Key features include:

• Private GKE clusters to keep workloads off the public internet.
• Workload Identity Federation to cut out the need for static passwords and keys.
• Cloud Armor to protect against web attacks and enforce strict traffic rules.

“Workload Identity Federation is a step forward in security practices because it eliminates a lot of issues that we had with private and public keys and making sure they are safe all the time and don’t expire,” says Niculescu.

The outcomes

The project proved to be a great success for Giles AI, as Insight was able to give them a fully operational Google Cloud environment in just 11 weeks.

“The speed of execution and the professionalism that we have experienced with Insight is nothing short of remarkable,” says Dhaliwal. “It has been so positive that we consider Insight part of our team. You don’t come across a solution provider like that often, and when you do, you want to hold onto them. If you’re looking for a long-term solution provider that can deliver exceptionally well, choose Insight.”

Quantifiable benefits

• Speed: Deployment times dropped from hours to minutes because changes are now deployed as code. “By removing manual configuration steps, we cut deployment times and reduced operational risk,” says Niculescu. “It also gives us a better overview of how things are going and the impact of changes we’ve made.”
• Financial clarity: Insight gave Giles AI a Google Cloud Calculator tied to their specific architecture. This lets Giles AI forecast spending accurately and plan for future scaling.
• Marketplace advantage: The new foundation makes Giles AI prepared for the Google Cloud Marketplace to let customers buy Giles AI software directly through Google Cloud, removing friction from the sales process.

Qualitative improvements

• Compliance ready: Compliance is now enforced by the architecture itself, not just policy documents. This creates a verifiable security stance that helps with SOC 2 audits and GDPR.
• Team autonomy: Through over-the-shoulder training, the Giles AI team learned by building. They now feel fully confident owning and expanding the system.
• Strategic relationship: The collaboration went deeper than just technology. “Insight believes in us, and we believe in Insight,” says Dhaliwal. “The highlight of this engagement was establishing our relationship built on mutual trust.”

Future outlook

Giles AI’s future outlook is focused on global expansion, scaling to meet the demands of the healthcare and life sciences sector, and upholding their mission to save time and ultimately save lives by accelerating vital research. The company has positioned themselves for this growth by building a scalable Google Cloud foundation, ensuring the solution can expand on demand and won’t require costly re-architecting in the future.

By becoming marketplace-ready and prioritizing compliance with GDPR and SOC 2 and future adoption of ISO 27001 and HIPAA, Giles AI has secured an easier, frictionless go-to-market strategy that is expected to open doors for their growth trajectory.

“We are particularly excited about what's next. In the next few months, Insight is onboarding the giles^® AI research assistant to the Google Cloud Marketplace, allowing us to reach our B2B market even faster through an established sales network,” says Rishi Wadhera, CEO at Giles AI. “Thereafter, we’re eager to continue our collaboration with Insight to further accelerate our technical development, ensuring Giles AI remains at the absolute forefront of healthcare and life sciences innovation.”