Your Garage
AI Smart Import
AI Smart Import lets you upload a CSV file and automatically classify each row as a service record, part, or tire. The AI reads your data, maps it to the right record type, and lets you review everything before importing. Navigate to a car's Maintenance → Service history then open the ··· menu in the toolbar and choose Smart import to get started.
Paid feature
AI Smart Import is available on Club, Pro, and Factory plans. Free-tier users will see an upgrade prompt.
How it works
The import follows four steps:
Upload CSV → AI Classification → Grouped Review → Import
Step 1: Upload your CSV
On any car's Maintenance → Service history, open the ··· menu and choose Smart import. Drag a CSV file onto the upload area or click to browse. The file is parsed in your browser -- column headers are detected automatically.
Any CSV layout works. The AI adapts to whatever columns your spreadsheet has, whether it came from a shop invoice, a personal log, or an export from another tool.
One file can hold up to 1,000 rows. A longer file is rejected before classification starts -- split it and import the pieces one at a time.
Step 2: AI classification
The AI reads each row and classifies it into one of these types:
| Type | What gets created |
|---|---|
| Service Record | A service history entry (oil change, inspection, repair, etc.) |
| Part | An item in your parts inventory |
| Tire | A tire in your tire inventory |
| Service + Part | A service record with a linked part (e.g., "brake pad replacement") |
| Service + Tire | A service record with a linked tire (e.g., "tire installation") |
The AI also maps free-text descriptions to the correct service type, part category, and tire specifications. A confidence score is assigned to each row.
Step 3: Review
Part manufacturers and tire brands/models offer the same catalogue suggestions as the inventory forms on desktop and mobile. Tire model search accepts aliases such as PS4S. You can still enter custom values; changing a tire brand clears its previous model.
After classification, the dialog shows your data grouped into collapsible sections -- one each for service records, parts, and tires. Each row displays:
- The extracted fields (name, date, cost, type, etc.)
- A confidence indicator (green for high, yellow for lower)
- A type dropdown to reclassify if the AI got it wrong
Click any Date cell to open a popover calendar -- pick a date without having to type it. Other fields like name, mileage, and cost are edited inline.
Linked records (service + part or service + tire) show a note indicating both records will be created together.
Rows the AI could not classify, or whose dates, amounts, or field combinations need correction, appear in an error section at the bottom. Review those rows before importing. For linked service and inventory records, put the shared charge on the service record so it is not counted twice.
Write mileage units explicitly when possible, such as 12,000 miles or 19,312 km. TrackOut converts supported units to the stored mileage and retains the source unit for review. Ambiguous values are flagged instead of being partially read as numbers.
Blank date or mileage? It carries over.
When forward-fill is enabled on your account, service-record rows with a blank Date or Mileage inherit the value from the most recent earlier row that has one -- so repeating columns can be left empty in your spreadsheet. You can still override any inherited value by clicking the cell. Forward-fill is rolling out gradually; if you're not seeing it yet, every cell will need a value as before.
Step 4: Import
Click Import to create all records at once. Service records are created first, then parts and tires are linked to them automatically.
CSV format
There is no required column format -- the AI handles whatever layout you have. That said, the following columns are recognized and improve accuracy:
| Column | Used for |
|---|---|
| name / description | Record name or description |
| date | Service or purchase date |
| mileage / odometer | Mileage at time of service |
| cost / price / total | Cost or purchase price |
| type / service_type | Service type hint (e.g., "oil change") |
| manufacturer / brand | Part manufacturer or tire brand |
| part_number | Part number |
| category | Part category |
| vendor / shop | Where the work was done or part was purchased |
| model | Tire model |
| size / width / aspect_ratio / diameter | Tire size specifications |
Click the Download template link in the upload step to get a sample CSV with recommended columns.
Tips
- Mixed spreadsheets work best -- A single CSV with parts, tires, and service records in different rows is handled automatically. No need to split your data first.
- Review low-confidence rows -- Yellow-highlighted rows may need manual correction. Check the type and fields before importing.
- Reclassify without re-uploading -- If the AI misidentifies a row, use the type dropdown to change it. Fields are remapped on the client without calling the AI again.
- Linked records save time -- When the AI detects "brake pad replacement" or "tire installation", it creates both the service record and the part/tire in one step.
- Common service phrasings are recognized -- Receipt wording like "Oil Service", "LOF", "brake check", or "PERFORM PORSCHE OIL SERVICE" maps to the canonical service type (Oil Change, Brake Inspection, etc.) automatically.