AI-first, ROI-driven

The work your team still does by hand.

Sales follow-up, onboarding, delivery, client queries, reporting. one7x rebuilds one at a time, against a number you agree first.

Where it costs you

Five places the work stops scaling.

Most founder-led service businesses lose the same five things. Not to competitors. To their own operations.

  • Sales follow-up Enquiries sit for days. Nobody knows which were answered, or what was promised.
  • Onboarding Every client starts from zero. The same documents chased by whoever is free.
  • Delivery Work runs across WhatsApp, email and a spreadsheet. Status sits in one head.
  • Customer queries The same ten questions every week. Junior staff cannot answer them, so you do.
  • Reporting The MIS arrives late and nobody trusts it. You find out when the client complains.

What one7x does

One practice, four ways in.

Start with the first if nothing is running yet. Come in at any of the others if you already know what you need.

Start here

Workflow automation and AI operations

The work your team still does by hand, rebuilt into the systems you already run. One bottleneck at a time, against a number agreed before anything is built.

  • Sales follow-up Every enquiry captured, answered and chased without you watching it.
  • Onboarding Every new client starts the same way, with the documents chased for you.
  • Delivery Every job shows its own status, updated as the work moves.
  • Customer queries Your team answers from your own documents. Only exceptions reach you.
  • Reporting The numbers come out of the work, not out of someone compiling them.
  • Product development A product your customers pay for, from the first idea to something live and in use.
    • Deciding what to build, and what to leave out
    • Something people can use before it is finished
    • Live, with real customers on it
    • Changed by what those customers actually do
  • AI product strategy What to build, what to leave alone, and whether AI belongs in it at all — settled before anyone builds.
    • Where AI pays for itself at your volumes
    • What to build first, and what it is worth
    • Buy, build or leave it
    • If one bottleneck is obvious and nothing is built yet, this is the wrong thing to buy first
    • A roadmap your own team can execute
  • Fractional product leadership Senior product leadership inside your team, part time — owning the decisions, not advising on them.
    • Owning the roadmap and the trade-offs behind it
    • Running the product team you already have
    • Deciding which parts of the product AI should touch
    • A way of deciding what gets built that outlasts the engagement

Where this fits

  • Project-based services
  • Recurring retainers
  • Field and site work
  • Document-heavy processes
  • Multi-branch operations

How it is built

  • Nothing new to buyBuilt into the tools you already pay for. Anything third-party is costed before you sign.
  • Your accountsYour logins, your data. If we stop working together you keep a working system.
  • Workflow automationThe routine steps run on n8n, so what was built stays readable and changeable by you.
  • Code and AI agentsWritten where a tool cannot reach — judgement steps, messy documents, anything bespoke.

Who this is for

You have outgrown how you run.

one7x works with Indian service businesses where the founder is still inside the operations.

  • The sizeTwenty to a hundred and fifty people. Revenue growing, headcount growing faster.
  • The setupSpreadsheets, email, WhatsApp, and two or three tools that do not talk to each other.
  • The pressureYou want the next thirty clients. You do not want the next fifteen hires.

Before anything is built

Every proposal starts from a baseline.

Nothing gets built until one7x can say what the problem is costing you, in a number you already look at.

Illustration — the shape of a proposal, not a client result

An illustration of how a proposal is laid out. It contains no figures and no client results.
What is measuredConservative ExpectedUpside
Cost removed
Capacity created
Revenue recovered
Founder hours returned

Your numbers go here, from your systems. Every one carries the assumption behind it, so you can argue with the assumption rather than the total.

How an engagement runs

It starts with a number.

Three steps. Each one has to earn the next, and any of them can end it.

  1. Free assessment Thirty minutes on how the work moves today. Three working days later you get it in writing: where your capacity is going, and what is worth fixing first. No system access. Free · 30 minutes, then 3 working days
  2. Operating cost review Two weeks measuring what your biggest bottleneck costs a year. You get the baseline, the payback period, and sometimes a recommendation not to build it. Under 6 hours of your time
  3. One workflow, rebuilt Six weeks. Built into the systems you already use, in your own accounts, against the target you chose. Something usable in your hands inside thirty days. One at a time · measured at the end

Whether one7x stays on afterwards is discussed at the end, not now.

Straight answers

What this is not.

So nobody spends a call finding out.

  • Not a strategy deckNo workshops, no assessments, no PDF at the end.
  • Not a body shopNo hourly billing, no bench, no swapping names.
  • Not a transformation programmeOne bottleneck at a time. Nothing that runs eighteen months.
  • Not regulatory adviceNo compliance, tax or legal advice at any point.

Before you ask

The questions that come up first.

Can I see examples of work you have done?

No. This is a new practice and there is no client work to show yet. Saying otherwise would be the first thing you could not check. The assessment is the substitute — a call and three working days of work on your own business, free, before anything is agreed. Judge how one7x thinks from that.

What does this cost, and what am I committing to?

The assessment is free. The review is a fixed fee agreed before it starts, and it can end there: if the numbers do not justify building anything, one7x says so in writing and does not quote for the work. The build is quoted against the baseline that review produced, so you are agreeing to a number you have already seen.

My team will not use it. What then?

That is the most common reason this fails, and it is a design problem rather than a training problem. The build starts from how your people already work. WhatsApp stays WhatsApp. Nothing new to open. If a step needs a new habit, it gets built into something they already use, or it does not get built.

Where does our data go?

Into your own accounts, on your own logins. one7x builds inside the tools you already pay for and does not keep your data anywhere else. Access during the review is read-only and named in writing. Where an AI model is used, the data it sees is named in the proposal before you sign.

What if it gives a client the wrong answer?

Nothing reaches a client unread unless you decide it should. The default is that the system drafts and a person presses send on anything customer-facing. Where you later want a step fully automatic, it is turned on one at a time, with a log of everything it sent.

What happens if we stop working with you?

You keep a working system. It runs in your accounts, on your logins, and one7x hands over in writing what it is and how it is put together. There is no licence to cancel and nothing switches off.

Do we need to fix our data first?

No. Most of this starts because the data is a mess. The review measures what is actually there, and the build works around what is missing rather than waiting for it to be fixed.

Get in touch

Book a call.

Thirty minutes. Tell us how the work moves today, and where it stops. If one7x is not the right answer, it will say so.

Prefer to send it in writing?