How Intelligems cut support answer time in half
Intelligems turned 1,500 support tickets into reusable knowledge, built a product roadmap that updates itself, and gave the company an AI Brain that gets smarter every night.
Start with Type
Intelligems is building the infrastructure for adaptive, intelligent commerce. Instead of static websites that treat everyone the same, its agents make real-time decisions on price, content, and experience for each visitor — based on what will maximize expected profit. Internally, the company has set an equally ambitious goal: make every employee proficient enough with AI to build and share automations that improve core business processes.
The team already had broad access to Claude. But useful work still lived in individual projects, required individual setup, and disappeared into disconnected chats. Intelligems needed a way to turn those one-off interactions into shared company infrastructure.
A small group of early builders used Type to create three systems the whole company could inherit: a support knowledge engine, a living PRD and roadmap, and a central AI Brain.
“If you start from a clean slate, I think it would be easier if we would all just use Type—at least for the commercial teams. It creates team spaces, you can collaborate much better, and you have more oversight and transparency into AI usage.”
Before AI was useful but the work didn't compound
Support reps worked from separate projects. Product knowledge was distributed across meetings, Slack, and Notion. Useful customer context could be synthesized in one conversation, then die in a chat or exported document.
Individual setup also made quality hard to inspect and improvement hard to systematize.
“If everyone runs individually, you have no control over the quality. You have no insight into the usage.”
Now there's one shared space to build, review, and improve
Type gave Intelligems a company-wide surface where people, knowledge, and automations could live together. Workflows run centrally, outputs stay accessible, and the next workflow can build on what came before it.
Product, engineering, support, and commercial teams can work from the same thread instead of passing screenshots, copied answers, and documents back and forth.
“It can short circuit so many review cycles. You just work in this one thread together.”
The support queue started teaching itself
Every new Pylon ticket is synthesized and compared with the existing knowledge base. Type either enriches a matching entry or creates a new one. The resulting database now contains roughly 1,500 entries.
When a rep needs an answer, Type checks prior tickets, searches technical documentation, and analyzes the customer's store and product journey when more context is required.
Max estimates that the time between a rep opening a ticket and reaching a usable answer has been cut roughly in half. The answers arrive with more company knowledge attached and require less iteration.
“It's probably halved, if not more, compared to the previous solution. The answers are much more detailed, thought through, and need less iteration to get to the right outcome.”
The workflow moves from known answers to deeper customer analysis.
Prior tickets
Match the question to a solved case in the 1,500-entry support knowledge base.
Product documentation
Search Intelligems' technical documentation for a verified answer.
Customer context
Analyze the customer's store and product journey when more context is needed.
Knowledge base
1,500 entries
Answer creation
Roughly 2× faster
Shared systems that support, plan, and learn
A support system that learns
New Pylon tickets are synthesized into a 1,500-entry knowledge base that combines prior resolutions, technical documentation, and customer context.
“The answers are much more detailed, thought through, and need less iteration to get to the right outcome.”
A living PRD and roadmap
Meetings, Slack conversations, and Notion documents are consolidated every night into a central PRD and roadmap that everyone can access.
“Our head of product created a live PRD that updates every night. Everyone has access to it across the company.”
A company brain that closes the loop
Teams submit research and decisions to a company AI Brain. Nightly skills organize the knowledge and alert the team when something looks inconsistent.
“Because we can run it centrally in Type, it has this reinforcing loop.”
Every use makes the next use better
Support tickets reveal documentation gaps. Those gaps improve the company's knowledge. Updated knowledge improves future answers. The living PRD and roadmap keep product context current, while shared threads keep reviews and decisions attached to the work.
Halving answer creation time points to roughly twice the handling capacity for comparable tickets once a rep begins work. The queue is still human-led, but the time inside each answer is already shrinking.
This leverage arrived before Type had spread across the whole company. A small group of early builders created systems that the broader team can use from day one.
“The other piece is creating a synthesis in a central place that everyone has access to, so it doesn't just die in someone's chat or in a document somewhere.”
Capture
Teams submit customer insights, decisions, research, and documents.
Synthesize
Nightly skills organize the content into shared knowledge categories.
Improve
Type flags inconsistencies and documentation gaps back to the team.
The next workflow starts smarter than the last.
Let Type set up your first space for you.
In just a few minutes you'll be rolling with your first agent.
Start a 2-week free trial“I would definitely start with Type, and then see if you need Claude for some teams.”
Max Klug, Intelligems