A CRM for small agencies where an AI agent tells you where each deal stands and what to send next.
ClientRaving
ProductravingCRM
Where and whenSri Lanka · Australia · 2025
PlatformWeb app
The screen, annotated by hand · scroll
the story so farthe agent writes the deal's history, so nobody re-reads the thread
the next question, already writtenfour prompts drawn from this deal, not from a template
the evidence is one tab awaylogs, calendar, documents, notes: what the agent read, so its answer can be checked
The story, in the order I work
Given · why
What was there
Pipeline kanban, a per-deal workspace with the RAVE agent summarising interactions and suggesting follow-ups, logs, calendar and documents in a side rail, and a rule-based funnel builder for automated follow-ups.
Planted · what I owned
What I took on
Placeholder until confirmed.
Shipped
What went out
A pilot, 2025.
Sales dashboardFunnel kanbanDeal workspaceAI suggestionsFunnel builderOnboardingAgent persona
A pilot. The surprise was the change management, not the technology. Younger users are curious and poke at the agent; experienced operators, given the right approach, get far more out of it, because they already know what a good answer looks like. Designing for the second group is where the value is.
Experienced operators get more out of AI. Design for them.
