Dodge AI Raises $2.65M to Automate SAP and Enterprise Software Maintenance

San Francisco-based enterprise AI startup Dodge AI has raised $2.65 million from Accel and Google's AI Futures Fund to automate the maintenance and troubleshooting of enterprise software systems, with an initial focus on platforms such as SAP, Kinaxis and Microsoft Dynamics.

The round also included Schema Ventures, New Build Ventures, Antler and angel investors from the SAP ecosystem.

Founded in 2025 by Rebhav Bharadwaj and Aditya Thakur, Dodge AI is building what it describes as an AI control plane for enterprise application maintenance. Its platform sits across complex business software environments to identify the root causes of incidents, recommend fixes and automate routine support work.

The company says its technology is already being used by more than a dozen enterprises, half of which are publicly listed.

Automating the work behind enterprise software

Large companies depend on enterprise software to run critical operations across finance, supply chains, inventory, manufacturing and other business functions.

SAP is particularly important in this environment, with companies often running highly customized installations that have evolved over many years.

Maintaining those systems can require large teams of consultants and IT specialists.

Dodge AI argues that the traditional model has created a dependency on system integrators and outsourced maintenance teams. When an incident occurs, consultants investigate the problem, make a change and close the ticket.

Over time, however, the knowledge behind those fixes can become fragmented across tickets, documentation, configurations and the people who worked on the system.

Customizations can accumulate, while technical debt becomes increasingly difficult to understand.

Dodge AI is attempting to address that problem with AI agents capable of understanding the context surrounding enterprise systems and taking action on maintenance tasks.

An AI control plane for enterprise applications

Dodge AI's platform is designed to operate across multiple enterprise applications rather than focusing on a single software system.

It connects with systems including SAP, Salesforce, Microsoft Dynamics, Kinaxis and Oracle JDE.

The company describes its technology as an AI control plane for enterprise application maintenance.

The platform looks across business processes, customizations, IT service-management systems and legacy configurations to identify the potential root cause of an incident.

It can then recommend or execute fixes depending on the workflow.

The company says its agents can also resolve L1 and L2 support tickets, covering routine first- and second-line enterprise application issues.

The objective is to reduce the amount of manual work required to keep enterprise systems operating while creating a structured understanding of why those systems behave differently across individual organizations.

Building an "exception intelligence" layer

A central part of Dodge AI's architecture is what the company calls an exception intelligence context graph.

Enterprise software is rarely configured exactly according to its original design.

Companies add custom rules, integrations and workflows to reflect their own operations.

For example, a particular warehouse might allocate inventory differently from the standard process. A pricing rule may override another rule, or a background job may run at a particular time because of a historical business requirement.

These exceptions can be difficult to document.

Dodge AI's context graph is designed to map those rules and relationships.

The company believes these undocumented exceptions can become an important source of context for AI agents operating within production environments.

Rather than treating every enterprise installation as a standard implementation, the system attempts to understand how each individual company's software actually works.

That context can then be used when investigating incidents or automating maintenance tasks.

From a warehouse incident to an automated fix

Dodge AI provided an example involving a warehouse where a truck could not be loaded because a Goods Receipt Note was repeatedly producing incorrect information.

According to the company, its software traced the problem across SAP, Kinaxis and internal warehouse software.

Dodge AI says the platform identified the underlying issue and delivered a fix within minutes.

The example illustrates the type of problem the company is targeting.

Enterprise incidents frequently cross multiple systems, meaning the application where an error becomes visible may not be the system that caused it.

A support team may therefore have to investigate several interconnected platforms before identifying the root cause.

Dodge AI is attempting to automate that detective work.

An example involving inventory planning

The company also described a case involving an enterprise whose inventory planning process had been moved to overnight because SAP would crash when the process ran in the morning.

Dodge AI says it modernized the process and made it 132 times faster.

According to the company, the change freed a team of 10 people who had been maintaining the process and improved order allocation time by eight hours.

These figures are company-reported examples of the platform's impact rather than independently verified performance metrics.

The broader goal is to show that maintenance can become more than simply keeping legacy systems running.

Dodge AI believes maintenance data and the knowledge generated from resolving incidents can become a foundation for future enterprise transformation.

More than a dozen enterprise customers

Dodge AI says it currently works with more than a dozen enterprises across incident management and process optimization.

Around half of those customers are publicly listed companies.

The platform currently handles hundreds of queries every hour across SAP, Kinaxis and Microsoft Dynamics environments, according to the company.

The customer base gives Dodge AI exposure to complex enterprise software environments where maintenance work can involve significant amounts of institutional knowledge.

The startup's challenge will be expanding that technology across different companies without requiring extensive manual configuration for every deployment.

Its context graph is designed to help address that issue by building a company-specific understanding of the rules and exceptions within each environment.

Competing for a much larger enterprise opportunity

Dodge AI is entering a market that already has several well-funded companies focused on enterprise software modernization.

Conduct has raised $60 million in Series A funding and is focused on understanding and mapping custom SAP code ahead of SAP's deadline for ending mainstream support for ECC software.

Tessera Labs has also raised $60 million, led by Andreessen Horowitz, to automate ERP migrations.

Freehand raised $75 million to build AI agents for supply-chain operations.

These companies are approaching different parts of the enterprise software lifecycle.

Dodge AI starts with maintenance and incident management, with modernization emerging as a potential longer-term opportunity.

That gives the company a different entry point into large enterprises.

Instead of approaching a business because it needs to migrate an ERP system, Dodge AI is targeting the continuous flow of incidents, support requests and operational problems that already exist within enterprise software environments.

The market opportunity

Dodge AI says the enterprise maintenance market represents a $600 billion opportunity.

However, third-party estimates cited in the source are considerably lower for the narrower application management services category.

Verified Market Reports estimated the application management services market at $15.4 billion in 2024, while Spherical Insights estimated it at $50.6 billion in 2025.

The wider application outsourcing market, which includes both building and operating software, was estimated at $131 billion in 2025 by Mordor Intelligence.

The difference highlights how Dodge AI's $600 billion figure represents a broader company-defined opportunity rather than a directly comparable third-party market estimate.

The startup's argument is that enterprise software maintenance represents a much larger pool of spending when outsourced consultants, system integrators, support teams and related operational work are considered.

Backing from Accel and Google's AI Futures Fund

The $2.65 million round was led by Accel and Google's AI Futures Fund.

Schema Ventures, New Build Ventures and Antler also participated, alongside angels with experience in the SAP ecosystem.

Dodge AI was one of five startups selected by Accel and Google's AI Futures Fund for the 2026 Atoms AI cohort.

The cohort was selected from more than 4,000 applications, according to the source.

Each selected company receives up to $2 million from the two investors, split equally, as well as up to $350,000 in Google compute credits.

For Dodge AI, the backing provides capital and access to infrastructure as it develops agents capable of working with complex enterprise applications.

Accel partner Prayank Swaroop said application maintenance is one of the largest categories in enterprise technology that remains relatively under-modernized.

Maintenance as the entry point

Dodge AI's larger ambition goes beyond resolving individual support tickets.

The company wants its agents to continuously understand enterprise systems, identify problems and help organizations reduce technical debt.

The underlying idea is that maintenance itself can become a source of intelligence.

Every incident contains information about how an organization's software operates, where its customizations are located and which business rules are important.

Capturing that information could make future automation easier.

For Dodge AI, the immediate challenge is proving that AI agents can safely operate inside production enterprise environments.

Its competitors are approaching the same broad problem from different directions, including ERP migration, custom-code analysis and supply-chain automation.

Dodge AI is betting that the daily maintenance queue is the right place to establish that relationship with enterprises.

With $2.65 million in new funding from Accel, Google's AI Futures Fund and other investors, the company now has capital to expand its platform and customer base.

If its agents can reliably diagnose and resolve complex enterprise incidents while preserving the context behind those fixes, maintenance could become an important entry point for AI-driven enterprise software transformation.