AI-Native SaaS · Startup idea
AI-native CRM: the idea you could build today, the repos already trying
An AI-native CRM doesn't replace Salesforce, it replaces the salesperson's spreadsheet. The opportunity is the long tail of one-to-twenty-rep teams who never wanted Salesforce in the first place.
Why now
Two things changed in 2025-2026. First, the model layer (Claude, GPT-class, Gemini, Llama) got cheap enough to draft every outbound email at $0.001 each. Second, MCP made it feasible to give an agent read/write access to a contact graph without a custom integration per ICP. That kills the gap between "CRM" and "sales copilot", they're the same product now.
The idea you could build today
A single-table contact graph plus an agent that does three jobs well: (1) summarize every email and call into one row of context, (2) draft the next-step message in the operator's voice, (3) chase replies on a cadence the operator can pause from a chat command. Skip pipelines, dashboards, forecasting, those are sequels, not the wedge.
Build stack
- ·Next.js + Postgres (Neon) for the contact graph
- ·MCP server exposing read/write tools to any host
- ·Embedding store (pgvector or Turbopuffer) for context recall
- ·Inngest or Vercel Queues for the chase cadence
- ·Stripe for billing, €19/seat is the indie ceiling
The three repos already trying
Pulled live from our current-period signal index.
+573%
14-day velocity Δ
8 contributors
+200%
14-day velocity Δ
6 contributors
Open source durable execution system. Write code that’s fault tolerant, durable, and simple.
Framework migration
+32%
14-day velocity Δ
285 contributors
Matched against the current-period startup signal panel (enterprise-saas, ai-ml). Rankings shift weekly as the underlying GitHub activity moves. Read the methodology.
The seed-round pattern hiding in the trendline
Watch for repos that hit ~250 contributors and 1.5x commit acceleration over a quarter while still describing themselves as "open-source CRM", that's the dual-loop OSS-to-seed pattern Highrise / EspoCRM / Mautic ran before. The repos that flip the README to mention "AI" or "agent" in the same period are the ones raising in the next 6 weeks.
Frequently asked
Why not just build on top of Salesforce / HubSpot?+
Because the customer who wants an AI-native CRM is the one who never signed up for Salesforce. The market is the founder, the recruiter, the agent, the consultant, single-operator pipelines that today live in Notion and Apple Notes. They don't want a CRM with an AI bolt-on; they want a chat window that remembers context.
How is this different from Attio or Folk?+
Attio and Folk are still pipeline-first, agent-second. The bet here is that the next-generation product is agent-first, the table view is a side effect of the conversation, not the surface area.
Is the market big enough?+
There are roughly 30M solopreneurs, agencies, and one-to-twenty-rep teams worldwide who run sales out of an inbox today. €19/seat at 1% penetration is a €70M ARR business, and that's before vertical add-ons.
Use the signal, not just the idea
Watch this idea live, every week.
The repos above re-rank automatically as commit velocity, contributor growth, and new-repo creation move. Want the data feed for this idea wired into your own stack? The MCP server exposes every signal as a tool any agent host can query.
Related ideas
Updated 2026-08-25. The framing is editorial; the “three repos already trying” slot is generated from the live signal panel. Anonymity rule: we name public GitHub orgs, never individual founders or stealth teams.