Monta launches AI NOC Agent to cut EV charger troubleshooting from hours to seconds

Monta rolls out AI-powered Network Operation Centre Agent for EV charging operations

Monta has introduced its AI-powered Network Operation Centre Agent (NOC Agent) to tighten uptime, shrink troubleshooting time, and scale operations across large charger fleets. The company says it is already deploying the feature across its platform.

For operations teams, the promise is simple: cut through logs and firmware histories in seconds, surface root causes with evidence, and push fixes faster. Anyone with a session ID can request insights through a straightforward interface, reducing the need for specialized tools or lengthy training.

What the NOC Agent does

• Diagnoses issues without manual log-hunting: Monta's AI analyzes session data, firmware states, and hardware behavior to pinpoint likely causes.
• Surfaces evidence: It shows what it found and why, so operators can trust the recommendation.
• Suggests next steps: From firmware actions to configuration checks, the Agent provides a practical playbook.
• Lowers support load: Monta reports that 75% of customer tickets are resolved by its AI before a human steps in.

Why operations teams should care

Scaling charger networks usually means adding people to keep uptime and SLAs on track. Monta positions the NOC Agent as a way to flip that equation-small teams running more chargers with tighter control, higher first-contact resolution, and shorter mean time to repair.

Reliability pressures are high. In markets like the United States, public funding programs expect strong uptime performance-for example, federal NEVI guidance targets 97% uptime for funded stations. See FHWA's summary of the final rule for context: FHWA NEVI Final Rule.

Early results (reported by Monta)

• 220,000 connected charge points and millions of monthly sessions feed the models, according to the company.
• One US operator saw a DC charger success rate jump from 31.2% to 98.3% after 25 seconds of NOC Agent analysis.
• Monta states that its AI models improve continuously with each charging session, firmware update, and user interaction.

How it fits into day-to-day operations

• Session-first triage: Start with a session ID, get a concise diagnosis and an action list in seconds.
• Evidence-led decisioning: Use the Agent's rationale to approve resets, firmware steps, or escalation without waiting on senior engineers.
• Faster escalation: When human support is needed, the handoff includes structured findings so cases move quickly.
• Continuous learning: As the Agent ingests more sessions and outcomes, playbooks get sharper and more autonomous.

Operator checklist to pilot the NOC Agent

• Define success: Pick clear KPIs-first-time fix rate, mean time to resolve, and uptime per site.
• Choose a controlled pilot: Start with a subset of DC/AC sites with known issue patterns and mixed hardware vendors.
• Instrument the process: Route all new incidents through the Agent first to capture baseline vs. post-pilot deltas.
• Set guardrails: Establish when to accept AI recommendations and when to escalate (safety, payments, grid faults).
• Update SOPs: Embed the Agent's steps into existing runbooks so every shift follows the same flow.
• Close the loop: Log outcomes (fixed, partial, wrong diagnosis) to strengthen future recommendations.
• Train the team: Run short operator drills on session lookup, interpreting evidence, and executing suggested fixes.
• Review weekly: Track KPIs, share wins and misses, and adjust thresholds for automation vs. human review.

What Monta says about the direction of AI in charging

Casper Rasmussen, CEO and Co-Founder of Monta, puts it plainly: "The industry is excited about AI but still figuring it out, and some actors fear AI is too complex to implement. At Monta, we're not experimenting - we're deploying. With our AI Agent and our data backbone, we're proving how AI makes networks smarter and easier to operate every day. This will transform the industry: small teams running thousands of chargers at higher quality than ever before. The time to prepare is now."

He adds: "Troubleshooting used to mean long nights with engineers digging through endless logs. With the NOC Agent, those hours turn into seconds. It's not just about fixing chargers faster - it's about redefining what operational excellence looks like in EV charging."

"Every AI use case we've built got 10x better with full-stack access. Monta has an unfair data advantage - and that's exactly why our AI delivers real results in production while others are still experimenting."

"Every industry has its inflection point where automation moves from optional to essential. In EV charging, that moment is now. Operators should no longer ask how many people they need to run a network - they should ask how small a team can run thousands of chargers with better quality than before. That is the future Monta is building."

What to watch next

• Autonomous incident resolution rates by fault category (connector, payment, firmware, network).
• Impact on spare parts usage and truck rolls per site.
• How well the Agent generalizes across mixed hardware vendors and firmware baselines.
• Operator trust: consistency between recommended actions and real-world outcomes.

If you're preparing your people for AI-augmented operations

Upskilling your team on AI-assisted workflows can shorten your rollout and reduce errors. For structured learning paths by role and skill, see Complete AI Training: Courses by Job.