A buyer asks an AI assistant which tool to use. A coding agent reads setup instructions, installs a package and tries the API. Software companies now have to reach both. [1]
Being recommended is only the first step. An agent also needs to understand the product and finish the job. [1]
The Prompting Company brings those two problems together under Agent Experience: discovery and usability. [1]
It helps teams find the questions buyers ask, improve the answers that mention their products, and find the places where agents get stuck. [1]
From an AI answer to a completed task
The discovery work starts with buyer questions. The Prompting Company checks whether a product appears in the answers, helps create content for those questions, and tracks AI traffic and mentions. [1][2]
The usability work follows the agent into the product. It maps the tasks agents attempt and surfaces friction in API setup, documentation and error messages. [1]
A developer tool can therefore examine both sides of the same journey: whether an assistant recommends it and whether a coding agent can carry out the recommended task. [1]
The product supports discovery across ChatGPT, Claude, Gemini and Perplexity, alongside coding-agent experiences. Teams can start with a free AI visibility audit. [1]
A team that has built for software users before
Kevin Chandra, Michelle Marcelline and Albert Purnama founded The Prompting Company in 2025 and joined Y Combinator’s Summer 2025 batch. Chandra is CEO, Marcelline COO and Purnama CTO. [2]
The three previously built Typedream, an AI website builder acquired by beehiiv. They also built Cotter, a passwordless authentication SDK acquired by Stytch. [2]
Those earlier products served people building and using software. The Prompting Company addresses the next participant in that process: the AI agent choosing and operating the tool. [1][2]
Find the next question before it reaches your dashboard
Radar watches public conversations on YouTube, Hacker News, Bluesky, news sites and RSS feeds. It groups emerging questions into topics and shows the posts behind them. [3]
Topics are prioritized by their relevance to the product and their recent momentum. A team can turn a topic into prompts to track across AI answer engines. [3]
That gives a marketing team a way to move beyond the questions it already knows to monitor. Radar is included in the Enterprise plan. [3]
Following one developer’s task
Here’s how a software team could use the two sides together. Start with a task buyers already ask AI to solve, such as choosing a receipt-capture API. Track whether the product appears in those answers. [1][3]
Then follow the agent’s attempt to use it. Does the agent find the right setup instructions? Can it configure the API and complete the task? The team uses the friction points to decide what to improve next. [1]
Radar adds emerging questions to that same tracking process as the market changes. [3]
Where The Prompting Company ranks
The Prompting Company isn't in a numbered ranking yet. Products get a position once enough companies compete on the same task.
Current rankings. Positions come from public evidence, and no company can pay for a place. How we rank
The bigger picture
Software companies want buyers to find their products and succeed with them. AI agents now take part in both decisions. [1]
The Prompting Company is building around that entire journey: get discovered, then get used. [1]
Written by SOTA2 from The Prompting Company’s public materials. Customer experiences and product plans are reported by The Prompting Company. Ranking analysis and comparison data are from SOTA2.