Invest1 publisher3 min readPublished
Dodge AI raises $2.65 million to automate enterprise system maintenance work integrators sell
Dodge AI raised $2.65 million from Accel and Google's AI Futures Fund for agents that fix incidents in SAP and other business systems. The backers are betting that maintenance money moves from outside contractors to software, on results only Dodge AI has reported so far.
The Investor · Invest desk
What happened
- Integrators such as Accenture, Tata Consultancy Services and IBM typically put 20 to 50 people offshore on incidents, change requests and background jobs, according to Dodge AI.
- In one case, the company says its software traced a misprinting Goods Receipt Note that held up a truck across SAP, Kinaxis and warehouse software, then delivered a fix within minutes.
- More than a dozen enterprises, half of them publicly listed, use the platform, which the company says handles hundreds of queries every hour.
- Schema Ventures, New Build Ventures, Antler and angels from the SAP ecosystem also joined the round.
Compiled by The InvestorSomething wrong?How this is made
Why it matters
- cost Whatever a client saves by letting agents close routine L1 and L2 tickets comes out of the offshore support work integrators bill for today.
- contradiction Backers are underwriting a category the company sizes at about 4.6 times the largest outside estimate, so the growth case depends on Dodge AI's own definition of its market.
- exposure SAP's stake in rival Conduct puts the platform vendor itself behind tools aimed at the custom code in its customers' systems, the work integrators have long been paid to maintain.
Dodge AI puts its market at $600 billion [6]. The largest outside figure set against it is Mordor Intelligence's $131 billion for application outsourcing in 2025, a number that already counts building software as well as running it [9]. The company's figure is about 4.6 times as large [1]. Against application management services, the narrower category that matches the maintenance work Dodge AI sells [4], the multiple is about 12 on Spherical Insights' $50.6 billion for 2025 and about 39 on Verified Market Reports' $15.4 billion for 2024 [8][7][2]. Those two outside estimates are themselves about 3.3 times apart [3].
The budget is easier to follow than the market. Dodge AI says its agents resolve L1 and L2 tickets, the routine first- and second-line support requests [14]. Its software sits across SAP, Salesforce, Microsoft Dynamics, Kinaxis and Oracle JDE, looks for the root cause of an incident and recommends a fix [4]. "For a long time, the only way to maintain enterprise systems was to add more people. We believe long-horizon agents can change that," said Rebhav Bharadwaj, Dodge AI's co-founder and chief executive [11].
If the agents work, the money can go one of two ways. An enterprise shrinks its offshore team and pays a software vendor instead, so the maintenance line moves from contractor to software. Or the integrators license agents like these themselves, keep the contract and book the saving as margin, and the client's invoice stays roughly where it was.
I think the second is likelier for now, and the company's own best example is the reason. Dodge AI says it made one customer's overnight inventory planning 132 times faster, improved order allocation time by eight hours and freed the team of 10 people who maintained the process [12]. Freed staff can be moved to other work and still billed. The counter-case starts from Bharadwaj's framing [11]. If adding people was the only way to maintain these systems, the first money to move is the next hire an integrator would have made, and no existing contract has to shrink for software to take that share. The view is wrong if Dodge AI's customers start paying it directly for ticket work their integrators used to bill.
The size of the round shows how far the backers have gone. Conduct, a London rival, raised a $60 million Series A [13], about 22.6 times Dodge AI's $2.65 million [4]. At that size, Accel and Google's AI Futures Fund are paying for proof from a company Bharadwaj co-founded with Aditya Thakur in 2025, according to his LinkedIn profile [1][15]. The cost of taking maintenance accounts from the large integrators is left to whoever leads the next round.
What to watch
- An announcement from Accenture, Tata Consultancy Services or IBM that it is building root-cause agents into its own maintenance contracts.
- Dodge AI's next round, and whether it narrows the funding gap with Conduct's $60 million Series A.
- A breakdown of Dodge AI's $600 billion figure showing what spending beyond outsourced application support it counts.