DataAgent Emerges From Stealth With $10M Pre-Seed to Build Self-Healing Cloud Infrastructure

DataAgent has emerged from stealth with $10 million in pre-seed funding as it launches an AI-native platform designed to move cloud infrastructure management beyond detecting problems and toward autonomously fixing them.

The round was led by MizMaa Ventures and Alicorn Venture Partners. DataAgent plans to use the capital to bring its platform to market and expand adoption in North America. Founded in January 2026, the Tel Aviv-based company currently employs 15 people and is led by co-founders Ishay Yaari, CEO, and Nati Shalom, CTO.

At the center of DataAgent’s approach is what it calls a remediation-first Digital Immune System for modern applications. Rather than acting primarily as another layer for collecting logs, generating alerts, or explaining incidents to engineers, the platform is designed to operate within Kubernetes and connected cloud infrastructure and take action when failures occur.

Reversing the traditional observability workflow

Modern observability platforms have become essential for understanding increasingly complicated applications, but the basic workflow remains largely reactive: a system detects a problem, an engineer receives an alert, someone investigates the cause, and a remediation is eventually applied.

DataAgent is attempting to reverse that sequence.

Its software reads live system state, topology, and configuration information where the infrastructure is running. When it encounters a failure it understands, the platform can apply a verified remediation first and conduct deeper root-cause analysis afterward. The goal is to restore service without waiting for a human engineer to work through logs and dashboards.

The company’s public website distills that philosophy into the phrase “Remediation-First, the rest is noise.”

The timing reflects a broader problem facing infrastructure teams. Grafana Labs’ 2025 Observability Survey found that complexity and overhead remain the most commonly cited observability concern, followed by signal-to-noise challenges and cost. Half of respondents now use Software-as-a-Service (SaaS) observability in some form, illustrating how much operational telemetry is increasingly processed through external platforms.

An autonomous SRE running inside the infrastructure

DataAgent describes its platform as an autonomous site reliability engineer (SRE), but its architecture is more constrained than that label might initially suggest.

The system builds a model of the customer’s topology, environment, code, and configurations instead of making decisions based on an isolated log entry or alert. During deployment, it identifies categories of failures it believes it can remediate reliably. It only acts autonomously on a fault category once that capability has been established, while unfamiliar or uncertain problems are routed to human operators.

Customers can also define guardrails around which actions the system is permitted to perform.

Those controls matter because autonomous infrastructure management carries considerably more risk than an AI copilot that merely recommends a command. An incorrect recommendation can be rejected by an engineer; an autonomous system with production privileges could potentially restart, scale, roll back, or modify workloads before a person intervenes.

DataAgent’s approach is therefore based on gradually expanding the classes of problems on which the system is allowed to act rather than granting broad autonomy from the outset.

Keeping more observability data inside the cluster

There is also an economic argument behind the architecture.

Traditional cloud observability often involves sending large quantities of logs, metrics, and traces to an external service where that information is ingested, indexed, stored, and analyzed. The cost can rise alongside infrastructure usage and telemetry volume.

DataAgent instead places its agents closer to the workloads themselves. Routine faults can be analyzed and remediated locally, with deeper inspection reserved for incidents that require it. The company argues this can reduce the amount of data that needs to travel through expensive ingestion and indexing pipelines while allowing its platform to sit alongside existing monitoring systems rather than requiring an immediate replacement.

This could become an important distinction as AI increases the amount of software being created and operated. More applications, services, containers, and AI agents ultimately generate more infrastructure events and telemetry. Simply collecting more data may become increasingly expensive unless software can determine which signals actually require attention.

Open source as a path to infrastructure trust

Another notable element is DataAgent’s decision to make its underlying agent open source. The agent can be deployed independently, while the company plans to monetize a SaaS layer for managing and orchestrating fleets of agents.

That model may help address one of the more difficult barriers facing autonomous operations software: companies are being asked to give an AI-powered system access to production infrastructure where mistakes can cause outages.

Being able to inspect and operate the agent locally gives infrastructure teams more visibility into what is running inside their environments, although the larger test will be whether enterprises become comfortable allowing those agents to take increasingly consequential actions.

Cloud orchestration veterans return for another infrastructure bet

Yaari and Shalom previously worked together at Cloudify, where Yaari served as Chief Revenue Officer and Shalom was founder and CTO. Dell acquired the cloud orchestration and infrastructure automation company in 2023 in a transaction that TechCrunch reported could have been worth as much as $100 million.

That background is particularly relevant to DataAgent. Cloudify focused on orchestrating infrastructure across Kubernetes, cloud platforms, and existing DevOps tooling. DataAgent moves further up the automation stack by attempting to give software the ability not only to orchestrate infrastructure but also to determine when that infrastructure needs to be changed.

The new funding will support DataAgent’s commercial launch and expansion, including additional go-to-market hiring in the United States and Israel.

From observability to self-healing infrastructure

The larger opportunity for DataAgent is tied to a transition already beginning across enterprise AI: moving from systems that explain what humans should do toward systems that can carry out the work themselves.

Infrastructure is a particularly consequential place for that transition. A large percentage of production incidents involve repeatable problems that experienced SRE teams already know how to resolve. If AI agents can safely absorb more of those routine tasks, observability could gradually evolve from a system for understanding infrastructure into a control layer that continuously maintains it.

DataAgent combines this production remediation loop with a pre-production system intended to identify and prevent failures before software is deployed. The company describes the resulting process as one in which lessons from incidents and prevented failures can progressively expand the system’s ability to operate autonomously.

The challenge will be demonstrating that autonomy can scale without introducing a new category of operational risk. Infrastructure environments differ enormously, unexpected failures are inevitable, and an automated remediation that works in one architecture may be dangerous in another.

If those safeguards prove reliable, however, the next generation of observability platforms may be judged less by how quickly they can tell engineers why something broke and more by how often engineers never need to become involved at all. DataAgent’s $10 million pre-seed round is an early bet on that shift.