We built an AI that tells you why your emails land in spam, not just that they did (launching InboxStack)

We're launching InboxStack, and I wanted to write this the way I actually experienced it, not the polished version.

The problem we kept running into

My co-founder spent years in enterprise deliverability — he built an internal dashboard called KARMA to track inbox placement across huge sending volumes. I kept hearing the same story from him: teams would see inbox placement drop, or get hit with a blocklist, and then... just guess. Was it a DKIM issue? A reputation problem on one IP? Something in the content? The existing tools would show that something broke, but never why.

That gap is basically the entire product now.

What we built

InboxStack is built around Brain, a causal AI engine that pulls signals from your ESP, DNS authentication records (SPF/DKIM/DMARC), Google Postmaster Tools, Microsoft SNDS, and 50+ blocklists — then actually traces the causal chain back to a root cause instead of dumping metrics on you.

The design principle we kept coming back to: monitoring isn't the bottleneck. Every competitor already has monitoring. Interpretation is the bottleneck. Nobody wants a chart that says "inbox placement -12%." They want "this dropped because of a DKIM misalignment introduced after your DNS change three weeks ago."

Getting to first users

We didn't want a multi-week onboarding slog, so the first step is just forwarding your email headers — that alone gives you an instant DNS/auth audit, no account, no migration. If you want ongoing monitoring and root-cause analysis, you connect your ESP and Postmaster accounts from there. Optional seed testing gives real inbox-vs-spam placement data across major providers if you want to go deeper.

Who it's for right now

We spent real time validating this instead of assuming it: digital agencies managing multiple client domains are shaping up to be our strongest early segment. Makes sense in hindsight — one agency account owner is dealing with N clients' worth of deliverability chaos simultaneously, so a tool that actually explains root cause (instead of just flagging problems across N dashboards) saves real hours, not just anxiety.

Pricing, if you're curious:

• Monitor — $49/mo

• Intelligence — $199/mo

• Agency — $399/mo

• Enterprise — custom

What I'd love feedback on

• If you run an agency or manage deliverability for multiple domains, does the "root cause, not just monitoring" framing actually land, or does it sound like something everyone claims?

• Anyone here dealt with Inboxbooster (YC W2017) back in the day? Curious what people think broke there, since we've thought a lot about not repeating it.

• Pricing — too cheap, too expensive, or about right for what it does?

Happy to answer anything about the tech, the causal graph approach, or how we validated the agency segment. Building in public from here on out.