What we built, and what changed.
Client names are left out on purpose, because the work and the data belong to the clients. Each case says what we built, how it works and what changed, in the tense that is true today.
Healthcare, telehealth & insurance
An online pharmacy and telehealth business in Ireland spent on ads every month for its brands and had no statistics of its own. We built the owner's own record of every consultation booking and a dashboard that shows ad spend per new customer and margin for each brand, independent of the ad platform's reports.
An insurance agency in the USA typed every contact twice, into an older CRM and into the enrollment platform. We moved its contacts, policies and notes into GoHighLevel and built a nightly update from the enrollment platform that linked records to contacts.
Construction & field service
A construction company's invoice and customer data sat in several separate tools, and its job management system did not link an invoice directly to a customer. We built that link and connected invoices to the company's tracking spreadsheet tool and project tool, with reminders and checks.
We built a connector that lets an AI assistant such as Claude read the data in a ServiceFusion account: jobs, customers, estimates, invoices and equipment, through thirteen read-only actions.
A construction company priced field jobs from price lists kept in spreadsheets. We built an estimating calculator for six service types of site work, with one shared set of pricing rules and one price list that the company edits in one place.
Real estate
We built the n8n system that researches US county deed records for a property research web app: page-by-page fetching with a saved position, keyword, location and demographic enrichment, an AI zoning analysis, and long jobs that run in self-starting batches, plus the app screens that launch and track them.
A US real estate agency's leads arrived through several forms and had to reach the CRM before anyone could follow up. We built the path from the forms to the CRM, a daily job that prepares each campaign batch for one approval click, and a specified follow-up process.
Sales & B2B operations
Systems we built for ourselves
A toolkit that lets our AI coding assistant list, read, create, debug and update automations across several servers. It drafts one from a plain-language description and checks it before it goes out.
An AI workspace with a database of client records and conversation history. Specialist assistants analyze project requests, draft proposals, write replies in our voice, estimate projects and prepare invoices.
It keeps a list of client conversations still waiting for our reply. An AI model judges whether a message needs an answer, and a card goes to our private chat with reminders until we answer or dismiss it.
A pipeline that collects public business signals from several sources, matches each company to one of eight customer types, scores the fit, finds a contact and drafts a personal message.
An open source memory server for AI coding assistants, public on GitHub. The assistant can record a decision, a bug fix or a convention, and later find it by meaning.
A toolkit our AI assistant uses to create servers, record existing ones, harden them to a written security baseline, deploy applications, and grade each server from A to F.
One shared room per project: a log of questions, answers, blockers and decisions in plain text. Each person connects with their own AI assistant and their own access key.
We built our own connector server that gives Claude, Cursor and ChatGPT authenticated access to a project-management tool, so tasks, comments and status updates can be made from the conversation.
We built our own AI sales agent platform. One agent with one knowledge base answers on a website chat widget, Telegram and WhatsApp, scores leads from the conversation and hands them to a person.
We built our own AI content system. Agents research, write, review and adapt a piece, a person approves it once, and publishers send it out to several channels.
Have something similar?
Tell us what is eating your team's time. We will scope the automation and send a fixed-price quote.