Tadit - Reliability intelligence for EV charging networks

Broken chargers don't always look broken. One shows as available but delivers nothing. A session never quite starts. Energy leaks out of the network while the dashboard stays green. There's no alarm and no ticket, so the loss only shows up as missing revenue and a driver who doesn't come back. Most of these faults never get investigated.

Tadit is reliability intelligence that sits on top of the CPMS you already run. It isn't another charging platform; it watches the one you have and does three things, continuously, for every charger you operate.

First, it catches every fault, including the silent ones. Thirteen fault types are detected out of the box, with nothing to configure. Beyond those, each charger learns its own baseline, so drift, a firmware version failing across a group, or sessions ending the wrong way all get flagged early. Faults are ranked by how busy the charger is, so you fix what drivers actually feel first.

Second, it finds the cause in minutes. Ask the AI diagnostic agent about any flagged charger and it investigates the way a good engineer would: it reads the charger's history, checks the manuals and service bulletins, rules out causes, and reaches a verdict with a confidence level and a full trail you can use in a warranty dispute. What used to be half a day of log-digging becomes a five-minute read, and every confirmed verdict feeds back into detection.

Third, it files the issue where your team already works. Each fault arrives as a single briefed ticket in Jira, ServiceNow or Slack, or as a webhook to whatever else you run, with history and severity attached. Playbooks apply the rules you set once, automatically, even at 2 a.m. with no one on call. A chronic fault comes through as one ticket instead of a hundred duplicates.

Tadit is also an MCP server. Connect Claude, ChatGPT or Gemini and ask your fleet anything in plain language. Answers come from live data, read-only, scoped to your workspace and secured with OAuth 2.1.

It also puts a number on what each fault costs, down to the kilowatt-hour, per station and per cause. You can fix the most expensive problems first and bring real figures to your ops review instead of anecdotes.

You can be live in days. It works with OCPP 1.6 and 2.0.1, needs no new hardware, and replaces nothing. It supports enterprise SSO, encryption at rest, cloud or self-hosted deployment, and your own AI keys. Your reliability data stays vendor-neutral, so you can compare hardware on the facts.