AI Automation & Business Tools in Kenya

AI automation that puts your business knowledge to work.

For consultancies and teams spending too much time finding, reviewing, and organising information.

Built around your business.

Useful automation begins with a repeated task: finding an answer in company documents, sorting incoming information, or preparing a first draft for review. We map that task first and decide where AI can help and where a person needs to remain in control.

I build AI features and workflow tools for business applications. Depending on the use case, that can mean a document assistant, structured extraction, automatic tagging, or a connection between existing tools. The aim is a specific, reviewable workflow with clear limits on what the system can access and do.

Illustration of connected business systems and data

What we can build together

A workflow assessment

A defined use case, representative examples, and a clear account of what a useful result looks like before implementation.

A focused AI tool

An assistant or automated step built around the agreed task, with output formats and review points that fit how the team works.

Appropriate data access

Decisions about source documents, permissions, retention, and provider use made with the team before business information enters the workflow.

Evaluation and handover

Checks against representative examples, documentation of known limitations, and guidance for reviewing output and monitoring ongoing usage costs.

From first conversation to launch

  1. 1. Choose one useful task

    Review the current workflow, the information involved, and the decisions people make. Agree on the boundary of the automation.

  2. 2. Test with representative examples

    Build a focused prototype and review its answers or extracted data against examples from the intended workflow, including cases where it should decline or ask for help.

  3. 3. Connect and monitor

    Integrate the agreed workflow, keep required review steps visible, and document access, costs, and the process for correcting failures.

Before we get started

Can an assistant answer questions from company documents?

Yes, when the documents are suitable for retrieval and the access rules are defined. A good implementation provides source references where possible and makes uncertainty visible so staff can check important answers.

Do we need AI for every automation?

No. Predictable rules and conventional software are often better for structured tasks. We use AI where interpreting language or unstructured information adds value to the workflow.

Can AI output be reviewed before it is used?

Yes. Review can be part of the workflow, especially for customer-facing content, sensitive information, or actions that change business records.

How are ongoing costs handled?

We identify the likely usage pattern and relevant provider costs during planning. Usage limits and monitoring can be included in the agreed implementation.

Tell me what you’re building.

Share the problem, your audience, and what a useful first version needs to do. I’ll help you define the next step.

Chat on WhatsApp

Andrew's AI assistant

Ask about work & services

A little help getting started

Hi, I'm Andrew's AI assistant. What are you working on?

Ask about his work, services, or your next project.