Sedai Secures $20M to Create the “Self-Driving Cloud”

Sedai, the company behind the world’s first "self-driving cloud," has raised $20 million in Series B funding, led by Atlantic Vantage Point (AVP) with participation from Norwest, Sierra Ventures, and Uncorrelated Ventures. The bigger story, however, is what this funding will fuel: a fundamental rethinking of cloud infrastructure management using autonomous AI.

Sedai’s executive team: Benji Thomas, CTO (left); Suresh Mathew, CEO (center); Vaneet Bhaskar, CRO (right). 

Rather than relying on human operators, dashboards, and alerts, Sedai deploys AI agents that make production-grade decisions in real time, autonomously scaling, healing, and optimizing cloud systems across AWS, Azure, and Google Cloud. The company has already executed over 25 million autonomous actions across its enterprise infrastructure, with zero recorded incidents.

A Control System for the Cloud

At the heart of Sedai’s platform is its Decision Engine. This engine serves as a coordination layer that orchestrates multiple AI agents, each targeting a distinct operational goal, such as minimizing latency, reducing cost, or maximizing uptime. 

These agents act not just on static thresholds but on continuously learned patterns. Using technologies like seasonality modeling, anomaly detection, causal inference, and deep reinforcement learning, Sedai adapts to changing traffic loads and evolving system behaviors.

The platform follows a four-stage cycle:

• Discover: Connects securely to cloud environments, identifies key signals (latency, errors, traffic, saturation), and ingests monitoring data
• Analyze: Builds behavioral baselines within 14 days by analyzing usage patterns
• Act: Takes autonomous actions, such as scaling, restarting, or reallocating resources, after passing built-in safety checks
• Learn: Evaluates the effectiveness of each action to continuously refine its optimization models

This feedback loop enables Sedai to function as an autonomous operations system—one that learns and improves over time, even as infrastructure scales.

LLM Self-Tuning and GPU Optimization

With its new funding, Sedai is preparing to expand into some of the most demanding corners of enterprise computing: large language models and GPU-intensive AI workloads.

Its self-tuning LLM capability will automatically analyze performance data, adjust system parameters, and reallocate memory and compute resources, all without human intervention. This ensures LLMs stay efficient and performant even as usage fluctuates unpredictably.

Sedai can help evaluate container infrastructure and choose the best type and number of instances, as well as how to group based on application behavior and reinforcement learning. 

In parallel, Sedai’s autonomous GPU optimization will intelligently manage high-cost GPU clusters using bin-packing algorithms and rapid scheduling. Inference throughput gains of 2.6–3.8× and training time reductions of up to 43% are expected based on early benchmarks.

Together, these features aim to transform enterprise AI infrastructure from a major cost center into a highly efficient, self-optimizing platform.

A New Operating Model for Cloud Teams

Sedai also supports a tiered operational model, including Datapilot (observe only), Copilot (human oversight), and Autopilot (fully autonomous). This allows teams to implement automation gradually, increasing their trust in the system over time.

In doing so, Sedai isn’t just replacing tools but, rather, it’s redefining the role of engineers in the cloud era. Instead of spending time on toil and tuning, teams can shift their focus to innovation and customer experience. As Sedai’s CEO, Suresh Mathew, put it: “Just like Waymo proved that self-driving cars are possible, Sedai proves that self-driving infrastructure is not only possible, it’s necessary.”

As cloud costs rise and talent constraints tighten, autonomous infrastructure may become a default operating model for forward-thinking enterprises. Sedai’s Series B marks more than a funding milestone. It’s a signal that AI-first operations are moving away from being purely experimental and becoming integral. With its platform already proven in mission-critical environments and new capabilities on the horizon, Sedai hopes to position itself as the control plane for the autonomous cloud era.

All images used courtesy of Sedai.