- Dodge AI has raised $2.65 million from Accel and Google’s AI Futures Fund.
- It says its AI agents fix incidents in SAP and other systems for more than a dozen enterprises.
- Its $600 billion market figure sits far above third-party estimates for outsourced app support.
A truck could not load at a warehouse because a Goods Receipt Note kept printing the wrong information. Dodge AI says its software traced the fault across SAP, Kinaxis, and internal warehouse software, and delivered a fix within minutes.
The San Francisco-based startup has raised $2.65 million from Accel and Google’s AI Futures Fund to do that kind of detective work across more of the software large companies cannot afford to break.
The round also included Schema Ventures, New Build Ventures, Antler, the early-stage investor that has backed more than 1,800 startups, and angels from the SAP ecosystem.
The startup calls its product an AI control plane for enterprise application maintenance: a layer that sits across systems such as SAP, Salesforce, Microsoft Dynamics, Kinaxis, and Oracle JDE, looks for the root cause of incidents, and recommends fixes.
“For a long time, the only way to maintain enterprise systems was to add more people. We believe long-horizon agents can change that. Dodge AI gives enterprises the ability to continuously improve and heal their mission critical systems so incidents can be resolved faster, technical debt can be understood, and the knowledge trapped inside maintenance work can become the foundation for future transformation,” said Rebhav Bharadwaj, co-founder and chief executive of Dodge AI.
The $600B claim
Verified Market Reports put the application management services market at $15.4 billion in 2024, and Spherical Insights valued it at $50.6 billion in 2025. The wider application outsourcing market, which includes building software as well as running it, was valued at $131 billion in 2025, according to Mordor Intelligence.
According to the company, enterprise maintenance has long run through large system integrators such as Accenture, Tata Consultancy Services, and IBM, which typically put 20 to 50 people offshore to handle incidents, change requests, background jobs, and other daily operational problems.
That model keeps systems running, Dodge AI argues, but fixes go undocumented, customisations pile up, and technical debt compounds. Companies grow more dependent on their maintenance partner because knowledge of how a system really works sits across tickets, consultants, configuration layers, and memory.
What the platform does, and what it has not shown
Dodge AI connects across business processes, customisations, IT service-management systems, and legacy configurations to pinpoint root causes and speed up the handling of incidents and change requests.
Underneath sits what it calls an exception intelligence context graph, which maps the rules that make each company’s system different: why one warehouse allocates inventory its own way, why one pricing rule overrides another, why a background job runs only at night.
The idea is that those undocumented exceptions become a source of truth for AI agents working inside production systems. The company says its agents also resolve L1 and L2 tickets, meaning routine first- and second-line support requests.
Bharadwaj and co-founder Aditya Thakur met at BITS Pilani, according to Bharadwaj’s LinkedIn profile. It says he co-founded Dodge AI in 2025 after deciding that maintaining enterprise software was a bigger, more structural problem than the browser agents he had been building for consumer brands.
Dodge AI says it works with more than a dozen enterprises, half of them publicly listed, on incident management and process optimisation across SAP, Kinaxis, and Microsoft Dynamics, and that its platform handles hundreds of queries every hour.
In a second example, one company had been running inventory planning overnight because SAP kept crashing when it ran in the morning. Dodge AI says it modernised the process and made it 132 times faster, freed the team of 10 people who maintained it, and improved order allocation time by eight hours.
Rivals with bigger cheques
Dodge AI is one of several startups going after the customisations buried inside large software estates, and most have far more money.
As Tech Funding News reported, Conduct raised a $60 million Series A, co-led by Index Ventures and ICONIQ, with SAP itself investing. The London startup maps custom code to the business logic it serves ahead of SAP ending mainstream support for its ECC software on December 31, 2027.
Tessera Labs raised $60 million led by Andreessen Horowitz on May 6 to automate ERP migrations, and TFN reported that Celonis and ServiceNow are also pushing into ERP automation. Freehand raised $75 million on July 29 for AI agents that take over supply-chain work large companies hand to outsourced teams.
Conduct and Tessera pitch themselves at understanding, changing, and migrating systems, according to their announcements. Dodge AI starts with the daily incident queue and says modernisation follows. It has raised a fraction of what each of those two collected: $2.65 million against $60 million.
Accel partner Prayank Swaroop, who has led the firm’s investments in Zetwerk and Slintel, said, “Application maintenance is one of the largest and least modernised categories in enterprise technology. Dodge AI is taking on the work that keeps the world’s most important systems running every day. By starting with maintenance, the team is building the context layer enterprises need before agents can safely operate in production.”
Accel is the Palo Alto-founded firm that backed Facebook, Spotify, and Atlassian early. It closed a $550 million India fund in August as part of a $3.5 billion global raise. Dodge AI was one of five startups that Accel and Google’s AI Futures Fund picked in March for the 2026 Atoms AI cohort, out of more than 4,000 applications. Each company receives up to $2 million from the two investors, split equally, plus up to $350,000 in Google compute credits.
Antler’s profile says roughly 70% of applicants were dismissed as “wrappers,” meaning AI features bolted onto existing software. Google launched the AI Futures Fund in May 2025 to back AI startups with early access to Google DeepMind models. It has also backed Emergent and Replit.
“Frontier models are incredibly capable, but enterprise adoption still requires significant hill climbing. Dodge AI captures the necessary rules and exceptions for this — while self healing your systems instead of an endless transformation project,” noted Thakur, co-founder and chief technology officer of Dodge AI.
For Dodge AI, the test is whether maintenance is a way in or a ceiling. Its better-funded rivals are chasing migrations with a hard deadline, while it is selling to teams whose immediate problem is the ticket queue.
If a $2.65 million startup can show public companies that agents fix live systems without adding to the technical debt it says it wants to remove, the consultant-led model looks weaker. If it cannot, larger rivals will get there first.
