Questions
The things people ask before working with me.
Including the ones that usually decide it: where your data goes, what I will not do, and what happens if the deployment does not work.
Ten questions
Straight answers.
What does Sonan actually do?
I deploy agents and automation on top of the tools a revenue team already uses, then keep them maintained. Deployment means a workflow automation, CRM and pipeline foundations, or a purpose-built agent for one specific job. The work is practical and narrow: one workflow, deployed and tested with the people who will use it.
How is this different from buying an AI tool?
Tools are inputs. Most teams already own more capability than they use, so we start by getting value out of the stack you have. Where a tool is genuinely missing, we say so. What you are buying from Sonan is the diagnosis, the deployment, and the adoption and maintenance work, not another subscription.
Do you send emails or update our CRM on our own?
Not by default. Every workflow starts with a person approving anything customer-facing and any significant CRM change. Once the output has earned it, you can move a workflow to full automation. That is your decision, made after we have walked through the risks, and it comes with automated checks before anything sends, volume limits, monitoring, and a stop switch someone on your side owns.
Where does our data go?
During discovery we work from public sources only. Implementation happens in systems you control, on your accounts. I do not need your customer database, call recordings, or anything financial, legal, or health related. If a scope would require more access than I can handle safely as a solo operator, I will say so before we start rather than after.
What does an engagement cost?
Three stages, each priced separately: a workflow review, a four-week deployment, and ongoing maintenance. Every stage is fixed scope with the price agreed before work begins, there is no hourly meter, and you can stop after any stage. The three prices are published here.
How long until we see something working?
The first useful result is usually weeks, not quarters. A deployment is built to produce something your team can inspect and use on real work, not a strategy document.
What if it does not work?
Then we stop and you keep the work product. If I conclude the workflow cannot be delivered safely inside these boundaries, I stop and refund the unearned balance. A failed deployment that costs a few weeks is a much better outcome than an agent nobody adopts.
Do you work with the tools we already use?
Yes, and the first one is usually a CRM. If there is no reliable CRM in place, that is the first thing we fix, because nothing downstream works without it. From there we work with what you already run, usually a CRM such as HubSpot, a call recorder and email, plus tools like Apollo for list building and data hygiene. Where a tool is genuinely missing, that can mean adding something like HubSpot Breeze, n8n, Claude or Hermes Agent, with the running cost shown before anything is built. I ran a HubSpot services business before this, configuring and supporting client instances and the automations behind them.
Who is this not for?
Teams that need a vendor to pass a SOC 2 review, companies looking for a fully autonomous outreach system, and anyone who wants a broad AI transformation programme rather than one working workflow. I will tell you on the first call if that is you.
What happens after a deployment ends?
You choose. Take the runbook and run it yourself, scope a second deployment, or walk away with the deliverables. There is no auto-renew and no default retainer. If you would rather the work be maintained for you, that is a separate month-to-month agreement you can end with thirty days notice, and that ongoing maintenance is the part most teams underestimate until something breaks.
Still deciding