14 projects, outside working hours

I built these myself.

In short

Fourteen AI agents and tools built by hand, outside working hours: sales assistants, candidate research, contract reading, Salesforce integrations. Most of them took three or four hours to build. They are not services for sale: they are the reason I know what it really costs to build what you are being quoted for.

When a vendor tells you that thing takes six months, I want to know whether it is true.

The fastest way to find out is to have built it. These are things I made in the evenings and at weekends, with my own money and with nobody handing me requirements. Some work, a couple I abandoned. Most of them took three or four hours.

They are not products: what you see in each video is the single use case, working. But once you have built it once, you know what it costs to build it again, anywhere. That is the difference between reading a quote and recognising the line that is inflated.

I do not sell them and I do not install them for you. They are here because when I look at a proposal and say "this line is inflated" or "the hard part is missing here", that judgement comes from this.

Sales

  1. Keith, a sales assistant Calls, books meetings, sends email and qualifies leads. Demo ↗
  2. Prometheia A team of agents working the sales opportunities nobody sees. Demo ↗
  3. Leads from the news Reads the day's news and pulls out companies worth contacting. Demo ↗
  4. Lead generation and enrichment Extracts contacts from news coverage and completes them with LinkedIn data. Demo ↗
  5. Sales automation The full pipeline, from first contact to proposal. Demo ↗

Salesforce

  1. Salesforce Canvas A different way of working inside Salesforce, without jumping between tabs. Demo ↗
  2. Salesforce and Chatbase An integration between the two, at zero licence cost. Demo ↗

People and process

  1. HR Research Agent Finds and analyses candidate profiles. Demo ↗
  2. HR People Finder Finds the right person for a specific task. Demo ↗
  3. Contract management Reads contracts and pulls out what matters. Demo ↗
  4. Customer service What actually changes when AI answers instead of a person. Demo ↗
  5. Data analysis From raw data to something you can decide on. Demo ↗

Other

  1. Mavis Turns an idea into a clean XML prompt for Lovable and similar tools. Demo ↗
  2. Property assistant Matching client to property in luxury real estate. Demo ↗

Why it matters to you.

When was the last time the person quoting you had already paid for that work out of their own pocket?