Video Summary - Full demo of Clarify CRM with the Co-Founders (July 2026)

Summary of the Clarify CRM demo (with co-founders)

Clarify is presented as an “autonomous AI CRM” built to solve the “scattered data” and data-entry tax problems that revenue teams face when managing deals, contacts, and activities across many tools. The founders emphasize that Clarify is designed as a warehouse-native “context/data layer” that unifies both structured CRM entities and rich unstructured activity data—so AI agents can work reliably inside the CRM.

Core product concept: an “autonomous” warehouse-native context layer

Clarify targets the root problem: data is scattered across many tools, and teams rely on downstream workflows in a warehouse rather than having a true unified record.

Clarify’s approach:

Centralizes customer data into one place (a data/context layer).
Stores not only relational/structured data (people, companies, deals, meetings/events), but also activity data as part of the same system.
Described as “warehouse native” and “warehouse of all your customer data,” managed by Clarify.

Why it’s “AI CRM”: the stack/philosophy (3 layers)

Andre frames Clarify’s AI CRM philosophy as requiring three stacked layers—missing any breaks the system:

Clean unified data underneath everything
- Includes structured, unstructured, and time-series/event data.
Conversation understanding over unstructured interactions
- AI understands emails, calls/meeting transcripts, and meetings.
Agents on top to do the work
- Agents perform real go-to-market actions/automation once the first two layers provide up-to-date context.

Key warning: agents must not run on stale data; otherwise automation amplifies mistakes.

Data layer capabilities: “true customer 360 view” + enrichment

The demo highlights that Clarify aims to keep CRM records automatically current:

Out-of-the-box sync with sources like email and calendar/meetings, so leads/meetings/deals are “automatically updated in one place.”
Unstructured capture + AI extraction from:
- Email bodies
- Call transcript data
- (And generally recorded meetings via an integrated call recorder and supported 3rd-party recorders)
AI-driven enrichment fills missing fields via an automated “waterfall” enrichment process (e.g., company details, contact info, location, firmographics such as company size).
A “good data in, good data out” model is explicitly stated.

Rep: conversational intelligence agent inside Clarify

A major product component is rep (Clarify’s conversational AI assistant / agent platform):

“Chat” is positioned as useful because rep has the full CRM context (history, records, deals, meeting summaries, etc.).
rep can:
- Capture and summarize calls
- Read email threads
- Extract decision makers and blockers
- Identify deal “must-have vs nice-to-have”
- Update fields and write summaries automatically
The system can also:
- Draft follow-up emails
- Create tasks from meeting outcomes

Agent automation: moving from chatbot to “do the work”

Clarify distinguishes between:

Chatting for answers, vs.
Specifying actions and letting agents execute them.

Agent behavior described:

Users instruct rep what to do (surface insights, complete actions, watch triggers).
Agents run on deterministic workflows + autonomous agent logic, triggered on schedules or events.
Proposed changes typically require approval (agents can suggest; they aim not to rewrite records silently unless configured/allowed).

Demoed interface surfaces and integrations

Desktop app experience

Desktop app “lives where you live,” reducing tab-switching.
Meeting prep is integrated:
- AI recommendations and summaries appear before calls.
- One-click follow-up email drafting.
Email and follow-up automation is a highlighted rep use case.

Slack integration as an interaction surface

rep can be used inside Slack, including:
- Asking questions
- Receiving actionable outputs and approval prompts
- Running work without needing to log into the CRM every day

Tool/agent harness + Anthropic ecosystem mention

Clarify integrates with Claude and was stated as the second publicly listed CRM in the Anthropic marketplace.
Clarify exposes “tools” to an internal agent harness so agents can operate and write back (and those tools can also be available in other agent-harness environments).

Connectors mentioned in the demo

Google and Microsoft calendar/email integrations (Google emphasized for sending emails and events).
Slack
Notes about integrations/tools used for the team’s internal GTM operations (e.g., Notion, Linear).
LinkedIn-related approaches:
- A custom object tracks LinkedIn connections (via export/upload; described as manual today).
- Customers are using LinkedIn automation tools externally (buyer beware for account bans).
- A Condo integration is mentioned as an inbox for LinkedIn DMs.

Calls/recorders

Clarify has an integrated call recorder and supports external recorders (Otter, Fireflies, Fathom mentioned).
Otter support described as via API or inbound webhooks; the integrated recorder and transcription data are positioned as central to agent workflows.

Data model (objects) and workflow-triggering design

The demo describes a flexible schema with both standard and custom objects:

Out-of-the-box: companies, people, deals, meetings, tasks
Custom objects and event-based modeling:
- Events, customer success/support process objects, “health snapshots,” partners
- Slack channels with customers
- Product usage data ingestion (referenced that an acquired company was Amplitude-related; Clarify can ingest usage data and trigger workflows)

Examples:

Events update fields (e.g., “last product interaction date”).
Workflows create deals, update CRM records, and trigger further automations.

Skills: codifying company voice and business context

Clarify describes skills as primitives that store contextual knowledge for better agent outputs:

Skills include organization-specific email/voice/tone, positioning, ICP/USP knowledge.
The system generates skills based on the customer’s go-to-market motion so agent drafts sound “like Clarify,” not generic AI output.

Agents marketplace + natural-language agent creation

The demo shows:

Creating agents via natural language prompts.
Agents library/marketplace with prebuilt agents such as:
- Inbox reply drafter
- End-of-day wrap-up
- Post-meeting follow-up

Agent creation includes:

Validating conflicts with existing agents (avoid overlapping logic/race conditions).
Using schema awareness (e.g., meeting schema for classifying sales calls).

Example agent: “weekly sales call digest” (scheduled Slack output)

A specific created agent example:

Weekly digest runs Friday at 4 p.m.
Inputs/logic:
- Pulls meeting/call data
- Synthesizes themes: call summaries, deal stage movement, objections, action items, next steps
Outputs:
- Formatted weekly update to a Slack channel
Includes privacy-aware behavior:
- Respects meeting sharing permissions/what meetings are visible to the user.

Built-in intelligent capabilities shipped out of the box

Mentioned capabilities include:

Deal suggestions
Deal summaries
Field updates
Meeting briefings
Calendar publishing of briefings
Task creation and other automations
Billing model ties to agentic work/inference/credit consumption (described later)

Prospecting features: Leadfinder + Campaigns

Leadfinder (AI search over large database)

Prospecting uses an integrated database of 200M+ records (people/companies).
AI surfaces relevant prospects (example: sales leaders evaluating CRM platforms).
rep generates queries and returns candidates quickly.
Candidate profiles include email/LinkedIn/company context and job history.
Users can import prospects into the system and then run outreach.

Campaigns (agent-assisted multi-step orchestration)

Campaigns can be generated from a prompt (e.g., webinar promotion with multiple emails).
Campaigns support:
- Creating campaign records and assets
- Enrolling signups into events
- Drafting email sequences using the organization’s voice/tone skills
Demoed an “omnichannel” direction:
- Customers request multi-channel campaigns (email + LinkedIn, and eventually SMS).
- Multi-channel automation is described as a top priority “in the works.”

Pricing model: charge for AI workload/inference, not seats

The founders position Clarify’s pricing:

Not charged per seat; described as “free CRM” in terms of unlimited users.
Not charging for data/call recording.
Charging for inference/AI workload (agentic usage).
Rationale: AI “software cost is going to zero,” but customers need paid value for agentic capabilities and workload.

Roadmap items highlighted

Clarify Signals (intent data) planned “next week or so”
- More intent data to improve forward-looking targeting (not just rear-view).
Multi-channel prospecting enhancements (email + LinkedIn, with more channels).
“Agentic reporting” / dashboarding updates
- rep can create charts/graphs; improvements expected in “the next few weeks.”
More reporting needs emphasized for larger customers (Series B+).

Q&A themes (key points from the customer questions)

Manager vs rep view: Clarify pushes actionable insights to managers (Slack/email notifications) rather than dashboards only.
Lead/contact verification: Leadfinder claims protections:
- verifies emails before sending campaigns to avoid domain reputation damage and bounces.
LinkedIn support:
- Third-party LinkedIn automation tools exist, but “buyer beware” due to bans/throttling connection requests.
- Clarify is evaluating native approaches and integrations (including Condo for LinkedIn DMs).
AI extraction accuracy/guardrails: Claims an agentic harness and guardrails to reduce hallucinations and prevent negative outcomes.
Dialers: mentions integrated dialers via API for logging transcript/activity data; support team can help set it up.
Zoom integration: described as enabling authentication to avoid getting stuck in the waiting room; transcripts are not pulled via Zoom integration alone (customers requested transcript ingestion).
Salesforce/HubSpot: currently early access; Clarify exploring bi-directional integration; users can request it.
Approval modes: AI can request permission before writing fields or update automatically with notifications.

Main speakers / sources

Patrick — Co-founder & CEO, Clarify (based in Seattle, Washington)
Andre — Co-founder & CTO, Clarify

Full demo of Clarify CRM with the Co-Founders (July 2026)

Key takeaways