Show HN: Recursive – AI support agents for small businesses

[alanmcintyre]

I'm a solo developer who's been writing code for 26 years, mostly consulting work in computational geometry and CAD software these days. A month or two ago I watched a YouTube video about using an AI agent as a resume assistant, and thought it sounded like a fun idea. More importantly, it sounded like something I could actually learn to build quickly.

Around the same time my longest-running client, a small company that makes a SolidWorks plugin, needed a better way to handle technical support. They're a three-person shop--they don't have a support team or an IT team. So I built them a chatbot: Flask backend, Anthropic API, and a hand-curated knowledge base built from their help files, a list of possible error messages I extracted from their code, and transcripts of their training videos.

It worked well enough that the owner was impressed, and that's when it clicked: every small software company has this exact problem. They're too small for Zendesk, too busy to answer the same questions repeatedly, and too niche for generic solutions.

So I turned it into a platform. You give me your documentation--whatever that happens to look like--I build your knowledgebase, stand up an agent on a subdomain of recursive.support, and you embed it on your site with a single iframe tag. No JavaScript SDK, no backend integration.

The part I'm most interested in technically is the self-improvement loop. The agent generates reports that review its conversations, identify gaps and trends, and propose specific knowledge base edits, presented as diffs the business owner can approve or reject individually. It sometimes even suggests fixes and enhancements I roll back into the product to make it better. The goal is that the agent gets measurably better over time without requiring the owner to manually curate anything.

There are five agents running right now: the original client's technical support, a documentation agent for an open-source project I maintain (neat-python), a founding documents Q&A at ask1787.com, the recursive.support visitor agent (dogfooding), and an AI resume agent that circles back to the video that started all of this.

I built ask1787.com, from idea to live deployment, in about nine hours, which is probably the best demo of what the platform can do at this point.

Tech stack: Python, SQLite, ChromaDB, Anthropic API (Claude Haiku & Sonnet), DigitalOcean droplets. Each customer gets an isolated deployment.

Happy to answer questions about the architecture, the self-improvement loop, or the experience of building a SaaS product as a side project while juggling 50 hours/week of client work.

On the actual business side, I have one paying customer--the original client whose support problem started all of this. Their contract covers most of the hosting costs for the entire platform, which means I can be patient about growth instead of desperate for it.

Fair warning: the product is live but early. So far my traffic consists mostly of people asking the agent to make them a hot anime girlfriend, so there's clearly some product-market fit work left to do.

[ai_tools_daily]

The small business angle is interesting. From what I've seen working with non-technical business owners, the biggest hurdle isn't the AI itself — it's knowing what to ask it to do. Most small biz owners I talk to in Australia haven't even tried ChatGPT because they assume it's 'not for them.' The ones who do try it usually start with something simple like drafting customer emails or writing job ads, and then the lightbulb moment hits. Support agents like this could work if the onboarding is dead simple.

[alanmcintyre]

Yep that's what I'm shooting for: the customer can just hand me a pile of documents, videos, etc., and I get them set up. They don't have to know anything about AI or web hosting if they don't want to (and a lot of small business owners don't). The tricky part now is getting it in front of people, but that's something I can work on gradually.