Sedai's AI Aims to Tame BigQuery Costs, Wins National Security Backing

SAN FRANCISCO, CA – April 22, 2026 – By Matthew Richardson

Sedai, a company branding itself as the creator of the “self-driving cloud,” today announced a two-pronged strategic move that underscores the growing importance of AI-driven automation in both commercial and government sectors. The company has launched an autonomous optimization solution for Google BigQuery, a powerful but notoriously costly data warehouse service. Simultaneously, it revealed a strategic investment and technology development agreement with In-Q-Tel (IQT), the influential not-for-profit venture capital firm that invests on behalf of the U.S. national security community.

The dual announcement highlights a significant trend: the push for autonomous systems to manage the spiraling complexity and cost of cloud infrastructure. For many organizations on Google Cloud, BigQuery represents a major operational expense, yet its consumption-based pricing model often leads to runaway spending that is difficult to control.

Taming the Beast of BigQuery Costs

Google BigQuery has become an indispensable tool for data analytics, but its power comes at a price. The service charges for data processed, meaning inefficient queries can lead to staggering, unexpected bills. Common missteps, such as using SELECT * to query all columns in a massive table or failing to partition data effectively, can cause costs to balloon. Industry reports show that companies can see their BigQuery bills double in a month from minor changes, while optimization efforts can slash those same costs by over 90%.

This is the problem Sedai aims to solve, not with more dashboards and alerts, but with autonomous action. The company’s new platform connects to a customer’s Google Cloud Platform (GCP) project and continuously analyzes how “slots”—the virtual CPUs that process BigQuery queries—are being consumed. It identifies where reserved capacity sits idle, where on-demand autoscaling is being used unnecessarily, and which workloads are inefficient.

"BigQuery slot mismanagement is one of the most expensive problems in a GCP organization, and one of the least visible," said Suresh Mathew, CEO of Sedai, in the company's press release. "Sedai eliminates that toil by autonomously managing slot allocation, so teams stop tuning by hand and start trusting their infrastructure."

Unlike many FinOps and cloud management tools that stop at providing recommendations, Sedai’s platform “closes the loop” by automatically executing changes. It analyzes demand patterns to right-size reservations and reallocates resources to prevent performance bottlenecks and reduce reliance on costly on-demand scaling, all while providing projections of cost and performance impact.

The Rise of the Self-Driving Cloud

Sedai's approach represents a broader shift in cloud management from human-driven optimization to autonomous systems. The current market is crowded with platforms offering cost visibility and recommendations, but they often leave the difficult and time-consuming work of implementation to already-overburdened engineering teams. This manual toil is precisely what “self-driving” platforms are designed to eliminate.

The core of Sedai's technology lies in patented machine learning models that learn an application's unique behavior, including traffic patterns, dependencies, and performance indicators. This intelligence allows the system to make proactive changes to cloud resources to meet performance targets (SLOs) while minimizing cost. The company claims this can be done without causing configuration drift from an organization's Infrastructure-as-Code (IaC) definitions, a common concern with automated tools.

While Sedai’s autonomous execution is a key differentiator, it is not entirely alone in this vision. Competitors are also emerging with a focus on autonomous BigQuery optimization, signaling a clear market demand for solutions that go beyond mere observation. The success of these platforms will hinge on their ability to earn the trust of engineers, who can be skeptical of “black box” systems that make changes without transparent, auditable logic.

A Strategic Bet on Autonomous Infrastructure

The partnership with In-Q-Tel provides a powerful validation of Sedai’s technology and its potential application in the most demanding environments. Founded by the CIA, IQT acts as a bridge between Silicon Valley innovation and the complex technological needs of the U.S. intelligence and defense communities. An investment from IQT is more than just capital; it is a signal that a technology is considered vital for national security.

"Now more than ever, the national security community depends on complex digital infrastructure to keep Americans safe," noted Katie Gray, Senior Partner at IQT. "IQT has chosen to invest in Sedai because the company offers a thoughtful and security‑focused approach to this challenge. Sedai's work to date demonstrates strong potential to help organizations manage critical infrastructure more efficiently."

This collaboration aligns with a major push within the U.S. government to leverage AI and automation to enhance operational efficiency, bolster cybersecurity, and maintain a technological edge. For government agencies managing vast and complex digital estates, both in the cloud and on-premises, the ability to autonomously ensure systems are both cost-effective and reliable is a strategic imperative.

As part of the agreement, Sedai plans to develop new capabilities to support government-specific requirements, a move that could significantly influence its product roadmap toward enhanced security and compliance. Mathew added, "This agreement with IQT is further proof that autonomous infrastructure has arrived."

Balancing Automation with Trust in Critical Systems

Despite the immense promise, the path to fully autonomous infrastructure is not without challenges, especially in critical sectors. Handing control over to AI agents introduces new security considerations. Cybersecurity organizations like OWASP have highlighted emerging risks, such as an AI agent with corporate credentials being manipulated through prompt injection attacks, potentially turning a cost-saving tool into a significant security vulnerability.

Furthermore, the adoption of these systems depends on building trust. If an autonomous system makes changes that engineers do not understand or cannot audit, they are likely to disable it. The most successful platforms will be those that balance autonomous action with transparency, providing clear explanations for their decisions and maintaining robust human oversight.

The investment from IQT suggests a growing confidence that these challenges can be managed. As autonomous systems move from managing commercial workloads to supporting critical national security functions, the frameworks for governance, security, and accountability will become just as important as the underlying algorithms. The journey toward a truly self-driving cloud is still in its early stages, but its direction is becoming increasingly clear.