# Dodge AI claims $2.65M can tame enterprise fires _Dodge AI raised $2.65M to push AI agents into SAP maintenance, but its $600B claim and autonomous fix story lack named customers and controls._ **Published:** 2026-09-29 **Source:** https://www.startuphub.ai/ai-news/funding-round/2026/dodge-ai-claims-2-65m-can-tame-enterprise-fires --- Dodge AI raised $2.65 million to automate enterprise maintenance, with Accel and [Google](https://www.startuphub.ai/ai-news/artificial-intelligence/2026/google-blackstone-form-ai-cloud-firm)'s venture arm leading the round. The San Francisco company says its platform acts as a control plane across SAP, Salesforce, Microsoft Dynamics, [Oracle](https://www.startuphub.ai/ai-news/artificial-intelligence/2026/openai-teams-up-with-oracle-cloud) JDE, Kinaxis and other systems of record. It connects to business processes, ERP customizations, ITSM systems and legacy configurations to pinpoint root causes, recommend fixes and capture the undocumented exceptions that define how each enterprise actually runs. That is a small check for a big claim. Co-founder and CEO Rebhav Bharadwaj said the only way to maintain these systems was to add more people, and that long-horizon agents can change it. Co-founder and CTO Aditya Thakur added that [frontier models](https://www.startuphub.ai/ai-news/artificial-intelligence/2026/openai-launches-partner-network) are capable but enterprise adoption still requires significant hill climbing, and that Dodge captures the rules and exceptions while self healing systems. Investor Prayank Swaroop at Accel called maintenance one of the largest and least modernized categories in enterprise tech. Dodge AI says it is already working with more than a dozen enterprises, half of them publicly listed, fielding hundreds of queries per hour. It cites two examples. A truck could not load at a warehouse because a Goods Receipt Note printed incorrect information, a fault it says it traced across SAP, Kinaxis and warehouse software in minutes. Another customer ran inventory planning overnight because SAP crashed in the morning, a process Dodge says it made 132x faster and freed a 10 person team while improving order allocation time by eight hours. Names, before and after metrics, and audit trails are missing. So is the security model for an agent that can recommend or apply fixes in production SAP. The announcement does not detail permissions, tenant isolation, change approval, rollback, or whether the agent acts remotely or requires local system access. For systems that hold pricing, inventory and financial postings, that is the real gating factor. The bet is that maintenance is the wedge to modernization. Enterprises want to move off brittle customizations but cannot risk breakage, and budgets are consumed by incidents. Dodge argues whoever understands the exceptions best gets to automate them, and turns maintenance exhaust into the operating manual for future agents. That logic holds if enterprises will let a seed stage control plane see and write to their core. The $600 billion figure describes the cost of not fixing it. This round does not yet prove they will. --- Original analysis from [startuphub.ai](https://www.startuphub.ai), the #1 AI startup directory.