2
2 Comments

I'm building a sales intelligence platform for AI agents (week 1 of building in public)

Hey IH πŸ‘‹

I'm Dave, and I've been quietly building Sales Stack β€” a pay-as-you-go sales intelligence platform built specifically for AI agents.

I've lurked here for a while. This is my first post.

What I'm building

One console that gives AI agents access to 5 data engines:

  1. Lead generation β€” search 300M+ profiles by seniority, industry, department

  2. Contact enrichment β€” partial email/name/linkedin url β†’ full profile

  3. Web scraping β€” clean markdown/HTML from any URL

  4. LinkedIn intel β€” profile data, company info, post engagement

  5. LinkedIn automation β€” visits, connection requests, messages

All JSON-first output. Works as an MCP server in Claude Code / Codex, or via REST API for tools like Clay.

Why I'm building it

I kept seeing the same pattern everywhere: founders and SDRs juggling 4–5 separate tools for prospecting, each with its own API key, its own subscription, its own stale database. And none of them were built for the way people actually work now β€” with AI agents doing the heavy lifting.

The existing tools were built for humans clicking buttons in dashboards. I wanted to build something an agent could just consume β€” ask a question in plain English, get structured JSON back, move on to the next step.

Where I'm at right now

  • All 5 engines are live and working

  • MCP server is functional (mcp add salestack https://trysalestack.com/mcp)

  • Real-time data (no pre-built lists)

  • Pay-as-you-go credits that never expire β€” free credits on signup, no card

  • Currently working on: onboarding flow, better docs, first real user feedback loops

  • Thinking about: pricing tiers (right now: $10 for 100 credits up to $150 for 2,500)

What's keeping me up at night

Honestly? Distribution. The product works. But I'm a builder, not a marketer. I'm not sure how to get this in front of the right people β€” SDRs, founders, people building AI agent workflows β€” without just blasting links into the void.

That's actually why I'm posting here. Not to promote (though free credits are available if anyone wants to try it). But because I've read enough IH posts to know that building in public beats building in isolation, and the feedback I get here will be worth more than any ad spend.

What I'd love from this community

  1. If you build AI agent workflows β€” what's the hardest part about getting real-world data into your agents? Is it finding the data, structuring it, paying for it?

  2. If you do outbound sales β€” what's the most painful part of your prospecting stack right now? Which tool do you hate the most?

  3. If you've launched on IH before β€” what worked? What didn't? What do you wish you'd done differently?

I'll be posting updates as I go β€” what I'm shipping, what's working, what's not, and the numbers behind it all.

Thanks for reading.

Dave

posted toAvatar for product sales_stack
sales_stack
  1. 1
    Your pay-as-you-go credits and the $10/$150 tier spread are exactly the kind of structure that gets hard to freeze once usage data arrives. I’m testing a narrow $299 one-time service for founders at that point: a read on three public competitor pricing pages, separating FACT/INFERENCE/UNKNOWN, with a specific recommendation. Sample: https://95gk4grdkw-byte.github.io/rival-brief/sample/ β€” are you dealing with that decision now, or would you still rather do the competitor work yourself?
  2. 1

    What stood out to me is that you're treating AI agents as the primary user instead of adapting software that was originally built for humans.

    That shifts the problem from "better sales intelligence" to "better infrastructure for autonomous workflows." If that becomes the mental model, I think the product is competing in a very different category than traditional prospecting tools.