Ask questions about your charging network in plain English and get instant answers
A conversational AI agent powered by your own data.
Your charging network generates thousands of data points every day: sessions, faults, payments, CDRs, driver profiles. Most of this data sits unused because querying it requires SQL knowledge or custom reports. The AI Agent makes your data accessible to anyone on your team - just ask a question.
Operations managers, finance teams, and business analysts who want instant access to their charging network data without needing technical skills.
Enable This ExtensionBrowse All MarketplaceEnable or disable per tenant. Pay only for what you use. All marketplace integrate seamlessly.
Ask anything: Which charger had the most faults last week? What was our total revenue in February? Get instant answers without SQL knowledge.
Powered by Private LLM running locally or on your infrastructure. Your charging data never leaves your environment.
Multi-turn conversations with full context. Ask a follow-up question and the AI remembers what you were discussing.
Create pie charts, bar charts, and dashboards via chat: Create a chart showing sessions by connector type this month.
Ask in any language, get answers in any language. The AI responds in the language you use - no configuration needed.
The AI only accesses your tenant data. No cross-tenant data leakage. Full audit trail of all queries.
How operators around the world can use this extension to solve real problems and grow their business.
A CPO operations team was spending 3 hours every Monday morning pulling data from multiple sources to create their weekly network health report. After enabling the AI Agent, the team simply asks: Generate our weekly network health report for last week. The AI queries all relevant data and produces a formatted report in under 5 minutes.
Talk to our team and see this extension in action with a personalised demo.
The AI can propose and execute operational actions — but with safety guardrails so the right humans stay in the loop.
Low-risk read-only and informational actions (get charger status, explain an error log) execute immediately without any approval step. Fast and frictionless for routine queries.
Actions like soft reset, hard reset, unlock connector, or dispatch a field ticket go to a pending queue for operator review. The AI explains its reasoning; the operator approves or rejects.
High-impact actions (set charger offline, update location tariff) require admin-level approval. Never executed automatically — full traceability from AI proposal to human decision to system action.
Every AI proposal, approval, rejection, and execution is logged with actor, timestamp, reason, and outcome. Immutable record for compliance and post-incident review.
Task-appropriate LLM routing: complex agent decisions and RCA use a frontier model; driver chat and log explanations use a local model (Ollama). Configurable per task type via environment.
When a charger goes into a fault state, the AI agent automatically analyses the preceding event sequence and proposes a root cause — with a confidence level and a suggested remediation action.