Case Studies
Solutions we’ve shipped.
Real AI systems, deployed to production, with the eval numbers and unit economics we actually run them by. Each case study names the tier it shipped at and what the customer pays today.
Logistics SaaS
Pro tier68%
tier-1 tickets resolved end-to-end
Multi-step support agent for a logistics SaaS — Pro tier
Built a chained support agent that classifies, retrieves, drafts, and hands off — wired to Zendesk, the customer's Postgres, and a hosted eval harness. Shipped on the Pro tier ($350/mo) in 8 working days. Now resolves 68% of tier-1 tickets end-to-end at $0.04 per resolution.
- median cost per resolved ticket
- $0.04
- classification accuracy on prod eval set
- 94%
D2C E-commerce
Pro tier+$9.80
AOV vs. holdout
RAG-powered merchandising assistant for a D2C apparel brand — Pro tier
Replaced a hand-maintained 'goes-well-with' spreadsheet with a RAG assistant the merchandising lead queries in natural language. Reads from Shopify + 26 months of clickstream + the brand's own copy guidelines. Pro tier ($400/mo). Lifted average order value $9.80 against a six-week holdout.
- cost per merchandiser query
- $0.018
- first production deploy
- 11 days
Financial Services
Custom tier92%
applications pre-decisioned with audit trail
Compliance-aligned underwriting assistant for a working-capital lender — Custom tier
Custom-tier build for a regulated lender that needed extraction reliable enough to feed a credit model and auditable enough to explain to examiners. Embedded pod owned model selection, retrieval, eval suite, and the SOC2-aligned audit ledger. Now pre-decisions 92% of applications with a fully cited reasoning trail per field.
- extraction accuracy on credit-model fields
- 99.4%
- median time-to-decision (was 6.1 days)
- 11 hrs