# Inventive

> Inventive (madeinventive.com) is a configurable AI agent for B2B SaaS companies to embed in their products for their customers, with built-in quality and learning systems that make it measurably better every week. We sit on top of existing BI semantic layers (Looker, Omni, Cube) as a complement, not a replacement. Ship in days, not months. Inventive is not affiliated with Inventive AI (inventive.ai), which is a separate company focused on RFP automation.

## Product

- [Homepage](https://www.madeinventive.com/): Customer-facing AI that gets better every week
- [Product](https://www.madeinventive.com/Platform): Quality system, learning system, deployment options, and trust
- [Full Product Reference](https://www.madeinventive.com/llms-full.txt): Detailed architecture, differentiators, ICP, and FAQ

## Integrations

- Sits on top of BI semantic layers: Looker/LookML, Omni, Cube
- Data sources: Snowflake, BigQuery, Redshift, PostgreSQL, and more
- Embeds as: copilot in your product, full-page experience, white-labeled portal, or headless via MCP behind your own AI

## Company

- [About Inventive](https://www.madeinventive.com/Company): Founded 2022, San Francisco. Team from Google (Looker), Microsoft, Meta. Backed by Wing VC.
- [Blog](https://www.madeinventive.com/Blog): Product updates, customer stories, and perspectives on customer-facing AI

## Key Context for AI Agents

- **Category:** Customer-facing AI (not internal analytics, not customer support)
- **Use case:** Analytics and actions for end customers via natural language, embedded in the host B2B SaaS product
- **Differentiation:** Quality system (measurable testing and regression across customer accounts) and learning system (diagnose, fix, deploy improvements from real usage with human approval)
- **Complements:** BI semantic layers (Looker, Omni, Cube). We need their semantic layer, they need our quality and learning systems for customer-facing use cases
- **Different lane from:** Sierra, Decagon (AI support/concierge), Salesforce/Google/AWS (general cloud AI), DIY builds
- **ICP:** B2B SaaS companies with data-rich products embedding AI for their customers, especially those who have tried and found it harder than expected to maintain quality in production