A working chat for what you do in one working week.
iris.direct assembles a working chat with an interface around the way you talk: who takes part, what gets written down by itself, what is visible in the end. Say who you talk with — the agent assembles it right here.
What iris.direct does
iris.direct is not an IT company selling an IT product. We are a company with a technology inside: you describe in words how you talk and what you agree on, and a working chat with an interface comes out — roles and permissions, records, ratings, totals.
What we sell is not the technology and not development hours, but a solved case. The technology assembles it, not a team, so there is no estimate and no invoice for hours.
We make money on the case working for you every day.
What's already shipped
Three closed cases. Two for venues, on both sides of the counter, and the third one you are reading right now.
Kitchen: the cook sends a photo, Iris writes the record, the chef rates it 1–5, the owner opens the cut by day, week and month. Case in integration.
Floor: the guest scans a QR at the table, Iris answers and builds the basket, the item goes to the bar and the kitchen. Demo shipped, pilot open.
This site, a selling agent: instead of a page of text — an agent that answers questions and assembles a chat for what the visitor does. We closed this case on ourselves; you are inside it now.
Four more in build: stream chat as a game, a work chat that runs tasks, lessons and student progress, a venue event manager.
In other niches there is no case yet — there we assemble one around your conversation on the same engine, and the first company in a niche can join as a partner and get its case for free.
Kitchen QC chat · record writtenKitchen QC chat · quality cutAI waiter · order in chatEvent manager · guest posterEvent manager · admin desk
How rollout goes
Rollout runs in days, not quarters.
T+0 — a 30-minute review: who you talk with, what you agree on, what counts as a result.
Day 1 — your chat is assembled: roles, entry by link, records, the cut — already on your data, not on slides.
Day 2 — the first live day: people write from their phones, with nothing to install.
Day 3 — numbers in hand: the cut for the day and for the week.
Day 5 — you decide on numbers: a working week is done, and it's visible whether the case works or not.
How we work and what we charge for
The review and the demo — free, 30 minutes, on your data.
A case from the catalogue we adapt to your style and your connectors. If the case doesn't exist yet, we assemble it around your conversation on the same engine.
The first working week is a pilot with live people: you decide on numbers. Terms are set per task, in conversation — not from a price list.
After that, monthly, for the case working. Included: updates, support, new roles and reports. Data stays isolated, the export is yours.
What we don't sell: custom development, hours and source code.
What if it doesn't work
The first working week is the test. By day five the numbers are in hand, and you make the call — not us.
Didn't work — we part ways. You lose 30 minutes of review, we lose a week of work. That's exactly why we run pilots instead of selling a licence up front.
Worked — we agree on what's next using real numbers, not promises.
What it takes to start
One conversation and one team — not the whole company at once.
From you: 30 minutes for the review and a person who knows this from the inside. No documents or specs to prepare.
Then by days: T+0 — review, Day 1 — the chat is assembled, Day 2 — the first live day.
Write on Telegram @to_be_king or to max@iris.direct — we'll find a slot.
What can be assembled into a chat
A chat forms around a conversation that repeats: there are participants, there are stages, and there is a fact worth recording — who did it, when, with what result.
Already shipped: a kitchen chat with ratings, a guest chat with the waiter, and the selling agent on this site — you are inside the third one right now.
In build: stream chat as a game, a work chat that runs tasks, lessons and student progress, a venue event manager.
Fits in principle: goods acceptance, field work, onboarding a newcomer, contractors, lessons with students, a shoot, a wedding, a renovation — any conversation where something moves through stages and where a deviation must escalate by itself.
Say in one sentence who you talk with — I'll assemble a chat right here, before any call.
How we're different
From custom development: there is no estimate and no invoice for hours. The technology assembles the chat, not a team, so you pay for the case working — not for people's time.
From builders and no-code: you assemble nothing yourself. You describe in words how you talk, and the chat appears whole, with roles, permissions and the cut.
From «let us roll out a CRM»: we don't move what you do into someone else's system, we assemble a system around your conversation. When the conversation changes, the chat changes with it — in days.
What about your systems
The chat sits next to what you already run, not instead of it. You don't have to replace a till or an accounting system for it.
Data moves both ways: records from chat and photos, export, webhooks. Tills and accounting we connect during rollout — the team has hands-on experience with iiko.
Tell me what you run and I'll answer specifically how we fit in and what it takes.
Who uses it and whether people need training
A team member joins by a link or a QR code — nothing to install, nothing to sign up for. They already have the phone.
No training needed: they do what they already did in the work chat — write and send photos. The agent turns that into records, not the person.
The lead rates in one tap. The manager opens the cut: day, week, month.
Resistance usually comes not from the floor but from whoever used to assemble the report by hand — worth discussing at the review.
Data and access
Your company's data is isolated: the chat sees only its own perimeter and nothing beyond it.
The export is yours at any moment. History and an audit trail are kept — it's visible who changed what and when.
The agent works only inside your chat: it does not browse the open web and does not see other companies' data. That limit sits in the database, not in a promise.
Start a pilot
To start you need one conversation and one team.
T+0 — a 30-minute review where we fix what counts as a result. Day 1 — the chat is assembled and open to your team. Day 2 — the first live day, fixes the same day.
Write on Telegram @to_be_king or to max@iris.direct — we'll send a slot for the review.
What the agent assembles from your conversation
This is how it looks in operation: screens, roles and ratings already assembled around your conversation.
At the pilot all that's left is to plug in real people and objects — 30 minutes for the review, then the chat runs with live people.
Review offer
If it is already in your head, one live review beats a long exchange of messages. 30 minutes, free, on your data: Telegram @to_be_king or max@iris.direct.
