Talk to your battery data.

Cliviq connects your battery test data to AI through MCP.
Ask questions in plain language, get interactive analytics instantly.

Try the demo → See how it works

claude.ai — connected to cliviq battery MCP

New conversation cliviq-battery · 6 tools

You

Show me the degradation of CS2-35.

✦

Assistant

CS2-35 shows steady capacity fade with a knee point at cycle 312. Here's the interactive analysis:

✓ get_degradation(cell_id="CS2-35")

You

Compare it with CS2-36.

✦

Assistant

Added CS2-36 to the comparison. CS2-36 degrades 18% slower — the knee point doesn't appear until cycle 410.

✓ compare_cells(["CS2-35","CS2-36"])

Ask a follow-up… ↵

Degradation dashboard Live

ZoomToggleCSV

CS2-35 CS2-36 Knee

Current SOH

82.3%

Knee point

Cyc 312

Predicted EOL

Cyc 580

How it works

Three steps to conversational battery analytics

Connect your data once. Ask questions forever. No dashboards to build, no scripts to maintain.

Step 01

Connect the MCP

Add Cliviq as an MCP server in Claude. One URL, one API key. Takes 30 seconds.

"url": "https://mcp.cliviq.com",
"token": "your-api-key"

Step 02

Ask questions

Query your battery data in plain language. Compare cells, analyze degradation, explore dQ/dV curves.

Compare capacity fade for all NMC622 cells tested at 45°C

Step 03

Get live artifacts

Interactive dashboards render instantly in Claude. Zoom, filter, toggle series, export — all in the chat.

→Interactive degradation chart

→dQ/dV peak tracking

→Export as CSV

Capabilities

Built for battery engineers

Every tool is designed around how battery teams actually think: chemistry → cell → cycle → step.

Cell-level analysis

Capacity fade curves, SOH trajectories, knee point detection, and end-of-life prediction. Compare across cells, chemistries, and test conditions.

Cycle-level deep dive

Voltage profiles, dQ/dV and dV/dQ curves with automatic peak detection and shift tracking. See how degradation mechanisms evolve cycle by cycle.

Tenant isolation

Your data is completely isolated. Each organization gets its own namespace, API key, and storage. Nothing is shared, ever.

Polars + Parquet

Raw data stored as columnar Parquet files. Analyzed with Polars for sub-second query performance even on million-cycle datasets.

Use cases

Who is this for

Any team that generates battery test data and wants faster answers.

Cell manufacturers

QC engineers compare production batches instantly. Spot anomalies in formation data before they ship.

Automotive OEMs

Validation teams query aging test data across suppliers and chemistries without writing a single script.

R&D labs

Researchers explore new cell designs conversationally. dQ/dV evolution, degradation fitting, cross-cell comparison — all from one interface.

Get started

Try it with real battery data

Our demo uses CALCE battery archive data. Connect the MCP to Claude, get a free API key, and start querying in under a minute.

Open the demo →