You have someone on your team who knows everything about your product. They know the edge cases, why the app only works in Google Chrome on an older device, and they fix can the reported bug faster than you've finished reading the customer's email. Yet that same person has no appetite for dealing with your customer directly, and you already know from experience that such conversations don't lead to great interactions either.
That isn't because your expert doesn't want to pull their weight. Something else entirely is going on. In this post we explain what's really happening, why the usual fixes grate, and how you can put your best technical people to work in support without it costing them energy and without the customer noticing a thing.
Put two programmers together and you'll hear an efficient conversation. Short sentences, no detours, straight to the cause. "Does it work in incognito? Ah, then it's the cache." Within that register, that's exactly the right tone.
Now put that same programmer in front of a customer and that very register lands wrong. The customer has a problem that's costing them time or money, and first they want the feeling that someone understands their frustration. They only want to be offered a solution after feeling they were heard. A conversation between technical people and a conversation with a customer are two completely different things. Your developer has been trained for years for communicating with developers, and not for talking to customers. Nor should they be.
The real brake isn't answering your customer, it's the stress that comes with it. Anyone typing a message to a customer knows that every word counts and lands directly. So comes the rereading, the rewriting, and these days often another pass through ChatGPT to make it sound more human. An answer that was in their head in ten seconds ends up taking ten minutes to write and send.
Companies try to solve this by not letting developers deal with customers directly, but by putting someone in between. That isn't just inefficient, it also raises the chance of mistakes, miscommunication and frustration.
Knowing the answer is something very different from phrasing the answer for the customer. So we've found a way to separate those two, so your expert gets to stay an expert in your product while Sarrai takes care of the communication.
That way the expert delivers the content and nothing more, in their own words, as short and as raw as they like. Someone else casts it in the right tone for the customer, adds the context, and makes sure it comes across as friendly and complete. The expert never has to think about how something sounds again. They only think about whether the information is right.
That's exactly the principle behind the Expert channel in Sarrai.
For most questions Sarrai has the answer straight away, based on your knowledge base. But sometimes Sarrai doesn't know, or doesn't know with 100% certainty. That's the moment the Expert channel kicks in.
Instead of guessing or leaving the customer waiting for a human, Sarrai summarises the customer's question, rephrases it clearly, and sends it as a short message to the right expert, roughly like a Slack message from a colleague. Does Sarrai already suspect an answer but isn't sure? Then Sarrai sends that draft along straight away and simply asks for validation. "I was thinking of answering this, is this correct?"
The expert can answer in their usual style. "Yes, but only on Windows 11." Short. Raw. No introduction, no pleasantries, no need for it to read smoothly. None of that matters, because that message doesn't go to the customer directly. Sarrai takes that raw input, rewrites it in your company's tone of voice, and rounds it out where needed with extra context Sarrai already has.
Meanwhile, Sarrai tells the customer it is just checking something with a colleague and asks for a little bit of patience, just like a human support operator would.
Now look back at the stress we described earlier. Thanks to the expert channel, that stress has completely disappeared for your expert.
Thanks to the expert channel, your expert is interrupted briefly. They know for certain their message won't go straight to the customer, so there's no stress about how customer-friendly their answer is. There's no need to reread, rewrite, or call in an AI of their own to proofread the reply.
What's left is the one part your expert is genuinely good at and actually enjoys: knowing what the right answer is. All other resistance to handling customer support is gone.
You don't have to cram everything into one channel. In Sarrai you can set up multiple Expert channels, each with their own members. For instance, you could have a channel for billing questions, one for the technical questions, one for a specific product line, etc.
Based on the topic of the question and the available channels, Sarrai decides for itself which channel to post in, and therefore which expert(s) to consult. You don't have to tag or forward anyone by hand. The right question automatically reaches the right person, without the other experts being disturbed unnecessarily.
Sarrai's knowledge base maintains itself, thanks to the conversations between Sarrai and your customers. The expert channel is no exception to this.
Every time Sarrai has to check something with an expert, it works out in the background what would need to change in the knowledge base to answer that same question automatically next time. Based on the answer the expert gave, Sarrai drafts a proposal for the necessary changes entirely on its own. All you have to do is review it and approve it.
That gives you a system that keeps getting smarter. Questions that need an expert today, Sarrai answers on its own tomorrow. That way, your best people will get interrupted a little less often each week, thanks to the continuous improvement of the knowledgebase.
You don't have to retrain your developers into support staff, and you don't have to push them into conversations that cost them energy. All you have to do is decouple their knowledge from the burden of phrasing it. That's exactly what the Expert channel does: it makes the knowledge of your best people available to everyone.
The customer talks to someone who comes across as warm and friendly and gives the right answer. Your expert delivers only the content, in ten seconds, without stress. And your knowledge base gets a little more complete with every question.
Create a free account and within a few minutes you'll see how many of your support questions Sarrai would already handle on its own today. Or request a demo and we'll look together at whether it fits your organisation.
A new colleague no longer opens the knowledge base — they just ping Tom, because that's faster and more reliable. That's how a knowledge base dies: not with a crash, but with silence. And that very gap is the most valuable signal you have.
Why does documenting processes so often fail, and what can we do about it? A handy insight into how to make your next documentation project actually succeed.
