❯ Question and answer
Where do I start with AI in my business?
Start with the bottleneck, not the tool. Map how your team works today, list where AI removes manual work, prioritize what pays back fastest, and only then build. That is Notkode's method, with 50+ projects delivered since 2020: diagnosis, opportunity map and implementation.
Updated on
What are the numbers, with sources?
- In June 2023, McKinsey estimated that generative AI, together with other technologies, could automate work activities that take up 60% to 70% of employees’ time. (Source: McKinsey, The economic potential of generative AI)
- Anthropic recommends starting with the simplest solution possible and only increasing complexity when needed. (Source: Anthropic, Building effective agents)
- OpenAI’s guide says to prioritize workflows with complex decision-making, hard-to-maintain rules and heavy reliance on unstructured data. (Source: OpenAI, A practical guide to building agents)
What is the first step?
Look at the operation before looking at tools. You do not need one more AI tool; you need to know what to apply in your business and why. The diagnosis shows how the team works today and where it loses time.
How do I know where AI pays back?
List the manual tasks that repeat and rank them by return. At Loss Prevention, the bottleneck was a sales team stuck in spreadsheets and email; the answer was 4 integrated tools, with leads flowing from the form to the CRM on their own.
How long until the first delivery?
An automation or AI agent is ready in 2 to 3 weeks on a normal timeline and within 1 week when urgent. A whole product takes longer: AutoAgentes went live in 8 weeks and generated more than R$ 70k in revenue.
Step by step
- 01
Diagnosis
Understand how the company works today, where the bottleneck is and what technology can solve.
- 02
Opportunity map
Show where AI delivers real results in the business and prioritize by what pays back fastest.
- 03
Implementation
Build the custom solution (a system, an agent or an automation), put it live and follow up until it delivers the agreed result.
Common mistakes
- Starting with an AI tool instead of the operation’s bottleneck.
- Trying to automate the whole company at once instead of one process at a time.
- Measuring success by tool usage instead of hours saved or revenue.
- Leaving the people who do the work every day out of the diagnosis.
Frequently asked questions
Do I need to hire an AI specialist to start?
No. The first step is understanding the operation; a partner can build, and your team learns to run what was delivered.
Which area usually comes first?
Customer service and sales: they repeat the most text tasks, such as answering customers and moving leads into the CRM.
Does my data need to be organized?
The data for the chosen process, yes. The rest of the company can wait its turn.
How long until the first result?
An automation or AI agent is ready in 2 to 3 weeks on a normal timeline and within 1 week when urgent.
Last verified on