Your next user may not be human.

Today, we’re announcing that The Prompting Company has raised $6.5M in seed funding led by Peak XV, with participation from Y Combinator, Standard Capital, Kearny Jackson, and Base10.

From AI discovery to Agent Experience

We started by helping companies understand how they show up in AI answers.

That was the first version of The Prompting Company: AI discovery.

But after working with more software, infrastructure, and developer tool companies, we saw the next problem.

Discovery was only half of the journey.

Even when AI systems knew about the right product, agents often failed to use it.

• They picked the wrong docs.

• They chose the wrong setup path.

• They missed the right API.

• They hallucinated implementation details.

• They got stuck on unclear errors.

• They failed halfway through the workflow.

That’s when we realized the bigger problem was Agent Experience.

The next customer journey does not end when AI mentions your product.

It ends when an agent successfully uses it.

How it works

1. Discovery

We find the exact questions your customers ask AI and test whether your product shows up across models.

We measure how AI systems describe your product, who they compare you against, what sources they cite, and where your product is missing.

2. Usability

We run agent workflow tests to see whether agents can actually use your product.

For devtools, infrastructure, and software products, this means testing whether agents can find the right docs, choose the right access path, use your MCP server, CLI, API, or SDK, and complete real tasks.

We look for the moments where agents fail: missing examples, unclear docs, broken workflows, API misconfiguration, bad error handling, missing context, or confusing setup paths.

Then we help you fix both through content, docs, and agent-ready pages.

Why this matters now

For the last twenty years, the internet was optimized around human users.

That is starting to change.

Agents are becoming the interface between people and software.

They will decide what tools to recommend, which docs to read, which API to call. 

This means every software company will need to ask a new set of questions:

• Can agents find us?

• Can agents understand us?

• Can agents choose the right path?

• Can agents use our product correctly?

• Can agents finish the workflow?

The companies that win the next decade will not just make something people want.

They will make something agents can find, understand, and use.

Why us

We’re Kevin, Albert, and Michelle.

We’ve been building together for more than a decade. We first went through Y Combinator in W20 with Typedream, which was later acquired by Beehiiv. Now we’re back building The Prompting Company.

Our belief is simple: as AI becomes the new interface to the internet, every company will need an Agent Experience layer.

What’s next

This funding gives us the resources to build the infrastructure for Agent Experience: large-scale AI discovery testing, agent workflow evaluation, analytics, content generation, citation monitoring, and optimization for AI-native traffic.

We’re especially excited to work with devtools, infrastructure companies, and software products with MCP servers, CLIs, APIs, SDKs, or technical docs that agents need to use.

If your next customer is an AI agent, your product needs to be built for agents too.

The Prompting Company helps you get discovered, chosen, and used by AI agents.