← All features
Planned
Free-Text Model Extraction
Parse unstructured text — call notes, customer emails, chat transcripts, voice-to-text output — and extract probable brand names, model numbers, and appliance types. Built for real-world operational inputs, not clean laboratory data.
POST /api/v1/extract (planned)
What it does
Extract likely brand and model number from messy free-text input like CSR notes or customer messages.
This feature is on our roadmap
Coming at launch — Spring 2026. Get early access to be notified.
How it works
- Your system sends raw text from CSR notes, customer messages, or transcripts
- ApplianceAPI scans for recognized brand names using a dictionary of 130+ brands
- Model-number-like patterns are extracted using appliance-specific heuristics
- Returns structured output: likely brand, model, appliance type, and confidence
Why it matters
- Automates extraction from CSR call notes: "whirlpool washer wtw5000dw making noise"
- Reduces manual data entry from customer emails and chat messages
- Enables voice-to-data workflows where technicians dictate instead of type
- Improves intake speed by auto-populating brand and model fields from free text
- Works with messy, incomplete inputs — not just clean, formatted model numbers
Who uses it
- CSR call note parsing
- Chat transcript enrichment
- Email intake automation
- Voice-to-data workflows
Current availability
Free-Text Model Extraction is planned. The initial version will use rule-based pattern matching against known brand dictionaries and appliance model formats. Join the waitlist to be notified.
Related features
Brand Normalization API
Map brand aliases, misspellings, and variants to canonical brand names. Clean up messy CRM and dispatch data.
Model Normalization
Fix messy brand and model strings from real-world CRM data before they hit your database.
Model Verifier
Validate model numbers before dispatch. Catch typos and invalid entries before they cause bad service calls.
Get early access
This feature is on our roadmap. Join the waitlist to be first in line.