Core Tools: Claude, Notion, Slack, Zapier + N8N, Notion, Halo ITSM, Monday.com Time Taken: 2 Weeks

As companies scale, internal support teams need to maintain more and more systems, then start to receive the same requests over, and over. "How do guests connect to the Wifi?", "where is our expense policy?", "I can't log in.."

These answers are (usually) written down somewhere, but everyone defaults to dropping messages in Slack or Teams because it's a more natural way of working. The challenge always comes down to friction.

There has to be a way for people to find solutions, get information easily, or triage a request to the right person, without disrupting that natural way of working.

I deployed AI powered knowledge bots directly into the support channels. The setup is simple and designed in a way that's easy to maintain after I handed it over.

Instead of feeding the AI generic data from the internet, I connected it directly into the company's internal knowledge docs and wiki. So when someone posts a question in Slack, Teams or email, the bot picks it up, searches internal documentation, and gives an immediate answer in a threaded reply.

Crucially, I built rules so that if the AI can't solve the issue (or if the requestor explicitly asks for help), the bot will triage the request. AI will create a ticket in the relevant support platform, populated with the context of the conversation, and route it to the team's queue, automatically tagged by priority.

Pixel-art conversation in a desert: a worried bird asks why its camera isn't working for a meeting, and a smiling robot suggests switching it off and on again\

By putting the bot in the place where people were already posting requests, it required zero behaviour change from anyone asking for help.

The impact was immediate and measurable across the organisation. By shifting to an AI assisted triage model, the company unlocked a ton of efficiencies:

Reduced Ticket Volume: More than 60% of queries are answered automatically by the bot, removing them from the support queue entirely.

Saved Time and Energy: Support teams are freed from sharing repetitive copy paste answers, allowing them to focus on solving more interesting problems.

Improved Knowledge Base: Because the wiki and process docs are now actually referenced every day, teams can spot gaps and update them easily.

The same template was rolled out across all internal departments, including HR, IT, and Legal + Commercial.

A solution like this is incredibly adaptable; budgets, team sizes, tech stack, everything's possible. I can customise precisely the workflow that works for you.