Services and pricing

Four steps, each with a flat fixed fee

Start with a free 30-minute call. Go further only if it makes sense for you. Every price is a flat fixed fee in US dollars. There are no hourly rates and no meter running.

At a glance

The whole staircase, prices included

Each step builds on the one before it. You can stop at any step.

  1. Step 1

    AI Data Health Check

    A 30-minute Zoom call with five questions about where your data stands today.

    Free
  2. Step 2

    2-Week Data Readiness Audit

    A scorecard and a 10 to 15 page executive readiness report.

    $3,500 to $5,000 flat fee
  3. Step 3

    Phase 2 Remediation

    4 weeks of hands-on work across three workstreams.

    From $8,000 by company size
  4. Step 4

    MCP / Data Governance Retainer

    Ongoing care, because data decays over time.

    $2,500 to $5,000 per month
Step 1 · Free

The free 30-minute AI Data Health Check

We meet on a Zoom video call. I ask you five discovery questions. They are simple to ask and surprisingly hard to answer.

There is no judgement. I start from where your data is today, and the five questions show where it stands, whether or not we go on to work together.

One of the five questions

“If an agent pulls your customer list today, which database is the absolute truth?”

Format
Zoom video call, 30 minutes
Hours
Monday to Thursday, 14:00 to 19:00 Cape Town, South Africa time
Price
Free
Step 2 · 2 weeks

The 2-Week Data Readiness Audit

A low-risk, high-value look at your data before you spend serious money on AI.

Over 2 weeks I assess your data against my four pillars of AI data readiness. You get a readiness scorecard for where you stand, and a written report for your leadership team.

I will sign an NDA before we start. You don’t have to give me direct access to your systems or chats, but you do need to supply the data from every system involved. Transparency is what makes the audit worth having.

Duration
2 weeks
Assessed against
My four pillars of AI data readiness
You receive
A readiness scorecard and a 10 to 15 page executive readiness report
Confidentiality
NDA signed upfront; no direct system access required
Price
$3,500 to $5,000 flat fee
Step 3 · 4 weeks

Phase 2 Remediation: the hands-on fix

4 weeks of work, in three workstreams, on turning scattered and incomplete data into one source of truth for your AI tools. The fee depends on the size of your company.

Flat fixed fees in US dollars for the 4-week engagement. No hourly rates.
Company sizePeopleFlat fee
Small Business20 to 50$8,000 to $12,000
Mid-Market50 to 250$15,000 to $25,000
Upper Mid-Market250 to 500$30,000 to $45,000+

The three workstreams

Workstream 1

Structured data clean-up

When I open a company’s systems I usually find duplicated records, data entered wrongly or in a hurry, and incomplete records. This workstream deals with your structured data so every client has one complete profile.

  • Mandatory field backfill. I identify and standardise missing high-priority fields such as account_tier, contract_status and primary_contact_id.
Workstream 2

Unstructured data architecture and RAG preparation

RAG (retrieval-augmented generation) is how an AI looks things up in your own documents before it answers. It only works if those documents are current and well organised.

  • Knowledge base deprecation and archival. Obsolete documents across Google Drive and Notion are retired and archived.
  • Document structuring and metadata tagging. Documents are shaped and labelled so they split cleanly into searchable pieces (vector chunking) and MCP tools can find them. MCP is the standard way AI agents connect to your systems.
  • OCR and scan processing. Image-based PDFs become searchable, structured text.
Workstream 3

Access control and MCP security guardrails

An AI agent can see whatever it is allowed to see. This workstream sets out what that should be.

  • User permission alignment. Role-based access control (RBAC) is enforced across shared drives and Slack, so people and agents only reach what their role needs.
  • PII redaction pipeline. Customer personal information (PII), card details and sensitive financial data are redacted.
  • Sandboxed testing environment. A separate space to test what agents write and do before anything goes live on MCP.

Working with your IT team. I’m very happy to work alongside your IT department. For remediation I would need one or two dedicated people from your side.

Step 4 · Monthly

The MCP / Data Governance Retainer

Clean data does not stay clean on its own. New records arrive, documents pile up and your database changes shape. Data decays over time, and the retainer is the monthly work of looking after your single source of truth.

$2,500 to $5,000 per month

Included each month

  • Monthly data quality audits
  • Tagging and indexing new unstructured documents added to Google Drive and Notion
  • Updating your MCP tools as your database schema evolves
Start at the bottom step

30 minutes, no cost, five questions

Book a Zoom call on Monday to Thursday, between 14:00 and 19:00 Cape Town, South Africa time. I’ll ask five questions and we’ll see where your data stands today.