AlertD, a startup delivering an agentic artificial intelligence site reliability and DevOps platform, launched today to give engineers highly contextual visibility across Amazon Web Services environments through an intuitive, conversational interface.

Founded in 2024 by Chief Executive Geoff Hendrey and Chief Operating Officer Freddy Mangum (pictured, left to right), the company was built around a simple guiding principle: “eliminating toil” for SREs and DevOps teams. Their goal is to exceed the limitations of legacy observability, log analysis and cloud operations alerting tools.

Hendrey and Mangum come from a background surfeit with tools built to maintain and monitor live software. Hendrey previously served as a principal architect at Splunk Inc. and chief architect at AppDynamics, part of Cisco Systems Inc. Mangum held leadership roles at Cisco and Fortinet Inc.

“In a nutshell, we’re AI for site reliability engineering,” Hendrey told SiliconANGLE in an interview. “After 10 years of building [observability tools], I realized it was one set of solutions creating the next set of problems — cognitive and telemetry overload.”

Modern SREs and DevOps teams, he added, are drowning in data and dashboards. Even sophisticated observability products pile on new layers of complexity and cost. AlertD’s vision is to offer an agentic AI teammate that can sift through that overload and deliver meaningful context in real time.

Contextual dashboards and collaborative intelligence

AlertD’s platform produces dashboards and answers. DevOps and SRE depend heavily on how information is visualized and presented; however, their needs change according to the types of data, the context that it’s in and where it comes from.

At its core, AlertD runs a daemon (the “D” in the company’s name), a background process that continuously indexes a customer’s cloud environment. This enables the system to fully understand available data sources and metrics, allowing its AI agents to dynamically create dashboards based on the actual state of a user’s AWS deployment, rather than relying on static documentation.

“What if dashboards were no longer a high-effort investment, but were instant?” Hendrey asked. “We translate an ordinary question you might ask into a data-gathering phase. Then the AI analyzes the dashboard itself for you.”

Hendrey likened traditional dashboards to “armor built from a previous fight,” handcrafted to reflect yesterday’s problems, but creating more problems when they don’t fit changes or their creator leaves. With AlertD, dashboards are ephemeral; “disposable free candy,” as he put it, automatically generated on demand and tied to contextual agents that understand the underlying environment.

On the front end, a conversational chatbot internally nicknamed Flappie (a nod to the company’s owl logo and Microsoft Clippy nostalgia) dispatches tasks to more than 150 specialized sub-agents. Each handles a specific type of inquiry, from database cost optimization to compliance checks or performance tuning.

“There’s still nothing like human knowledge,” Mangum said. “We have a strong capability that allows for human collaboration so that tribal knowledge is not lost into some script on someone’s desktop.”

Each workflow link can be shared internally, allowing engineers or auditors to rerun compliance checks and diagnostics on demand. Users can also build their own custom agents using the same agent-building kit AlertD uses internally and share them on an Agent Exchange, a planned marketplace for reusable, intelligent agents tailored to DevOps tasks.

“We’re able to add these into our product as agents,” Hendrey said. “We don’t need a whole new product stack to add that security use case or that cost-savings use case. We’ve actually built, very quickly, using the product, use-case-specific agents that can then be shared on an exchange.”

AWS focus, early traction and the path ahead

AlertD’s initial focus is deep integration with Amazon Web Services, but its architecture is cloud- and large language model-agnostic, designed to extend easily to other ecosystems. Users can bring their own software delivery lifecycle tools and LLM providers.

The company quietly raised $3 million in seed funding from True Ventures in October 2024.

“Earning that funding round is not a right; it’s a privilege,” said Mangum. “We just wanted to keep our nose to the grindstone and work closely with our design partners.”

From that development, AlertD is already delivering measurable value for beta customers. Metrics include a 10% cost savings for one unnamed enterprise customer through a cost optimization agent and a DevOps engineer who spent eight hours building a script for a query, then watched AlertD provide an answer in 37 seconds.

“We’re an agentic teammate that gathers context and answers questions quickly across a variety of domains,” said Hendrey. “Our main goal is to reduce toil.”

Images: AlertD