Doris
Doris is a semantic layer for go-to-market teams that reads every call, email, and document, then links each piece of information to the deal it belongs to. It maintains one typed, queryable record of what is true about each deal, and its AI "Sidekicks" use that record to draft briefs, surface winning plays from similar past deals, and automate follow-ups. The platform includes a notetaker that joins calls on Zoom, Meet, and Teams, and exposes its ontology to AI agents via a remote MCP server.
- Sales Technology
- AI Agents
- Artificial Intelligence
- Software Only
- Enterprise Software
Funding
Founders
Founder details are not available yet.
Product
Problem
Critical deal intelligence is scattered across call recordings, email threads, documents, and the memories of sales representatives. When a rep leaves, the knowledge of why deals closed or what buyers wanted leaves with them, forcing the company to relearn the same lessons on every new deal. This fragmentation prevents teams from consistently applying winning strategies and understanding the true state of their pipeline.
Solution
Doris is a semantic layer that sits underneath the go-to-market stack, reading every call, email, and document, then linking each piece of information to the deal it belongs to. It maintains one typed, queryable record of what is true about each deal, including buyer wants, objections, champions, and competitive dynamics. The platform connects each deal to similar past opportunities, surfacing what convinced comparable buyers before the next call. Doris also serves its entire ontology as a remote Model Context Protocol (MCP) server, allowing AI agents like Claude, Cursor, and ChatGPT to query the pipeline in a single call. The product includes a notetaker that joins calls on Zoom, Meet, and Teams, a sales agent that answers pipeline questions, and a flows builder that automates described work.
Target Audience
Primary customers are sales operations leaders, revenue operations teams, and sales representatives at B2B companies that rely on consultative, multi-stakeholder deal cycles and need to preserve and apply institutional deal knowledge.
Features
- Automated notetaker that joins every call on Zoom, Microsoft Teams, and Google Meet, capturing and transcribing conversations without any rep input
- Ontology engine that resolves fragments from multiple tools into one typed, queryable record per deal, including people, meetings, commitments, objections, competitors, and stakeholders
- Fact-level provenance, where each fact taken from a call links back to the exact passage it came from
- Deal similarity matching that links current opportunities to closed-won deals with shared buyer wants, surfacing winning plays, objections overcome, and buyer-accepted responses before calls
- Remote MCP server integration that exposes the full ontology to AI agents like Claude, Cursor, and ChatGPT for natural-language pipeline queries and write-back with the same guardrails as the app
- Flows builder that creates a typed graph of triggers, conditions, and actions to automate follow-ups, CRM updates, and brief drafts across independent branches
- Human-in-the-loop controls ensuring nothing reaches a buyer unseen and every generated message requires approval before sending
- Weekly usage analytics per person, showing how many Sidekick runs, approvals, and actions each user handled