Skip to content
ZerekAI

Process

How we work

No endless phases or “transformation roadmaps”. Five clear steps: each one shows what we do and what we need from you, so no time is wasted on back-and-forth.

01

Assessment: 30 minutes

We review one specific process: how many inquiries per day, how much of an employee's time they take, where leads get lost. We estimate the impact in hours and dollars, not in vague words about “artificial intelligence”.

we need from you: Half an hour of your time and an honest answer to “what hurts the most right now”.

02

2-page spec

We fix the channels (WhatsApp, CRM, 1C…), the scenarios, what the assistant does on its own and what it hands to a person, and 3 success criteria for accepting the work. The price is fixed here and doesn't change.

we need from you: Access to your current materials: price lists, FAQs, policies. This is what the assistant will learn from.

03

Build: 1–2 weeks

We build the assistant on servers in your country or region and connect the CRM and messengers. First we test on 10% of real inquiries, so errors are found before full launch, not after.

we need from you: One person on your side who answers process questions quickly. Delays in access shift the timeline.

04

Launch and training

We switch on 100% of inquiries and show your team how to read conversations and when to step in. We record the “before” and “after” numbers: they form the basis of a case study if you agree to share them.

we need from you: 20–30 minutes for a short training session with the team that will receive escalations from the assistant.

05

Support

Once a month we prepare a metrics report (how many conversations were closed without a person, where escalations happen most), update the knowledge base for new prices and services, and monitor errors. Incident response is within one business day.

we need from you: Send us changes to prices and services when they happen. Otherwise the assistant answers based on outdated data.

·Honestly

What we deliberately don't promise

  • ✕Custom ML models, computer vision or predictive analytics at the start. If you don't have accumulated data yet, that's a premature use of budget.
  • ✕A promise of “AI transformation of the entire company”. We take one process at a time, because that's how it can be measured.
  • ✕A 24/7 SLA with guaranteed replies within minutes. On a retainer we plan for a response within one business day, not round-the-clock duty.
  • ✕Replacing your CRM or 1C. We integrate with what you already have instead of selling migration for its own sake.

Let's start with an assessment: 30 minutes, free

Message us on WhatsApp →