AI implementation for small and medium-sized businesses
AI in one specific place in your company. Not an “AI strategy”.
I put AI where someone does something by hand, over and over: the Monday report, replying to emails, sorting tickets, questions about company documents. Connected to your data, with permissions set so AI sees only what it should.
- A developer, not a slide deck
- A prototype in 1–2 weeks on your data
- A separate, read-only account for AI
- A reply within one business day
Does this sound familiar?
- Someone assembles the same report from several sources every week.
- An inbox of enquiries that are mostly the same questions.
- Staff keep asking the same things: procedures, price lists, contract clauses.
- You bought an AI tool and do not know how to connect your data to it.
- You worry that customer data will end up where it should not.
What I implement
Recurring reports
Data from your CRM, spreadsheets and databases pulls itself, AI adds the commentary, every figure leads back to its source. Delivered by email or to the team channel.
Email and documents
Sorting and summarising email, draft replies for approval, an assistant that answers from your procedures and contracts and cites the source.
Chat on your website or shop
Answers the simple questions, collects contact details, hands anything harder to your CRM together with the transcript.
AI connected to your systems
Claude, ChatGPT or your own agent connected to your CRM, Drive and database through MCP or an API. Separate account, minimal permissions, a log.
Automations with AI inside
A workflow where AI does one step: classifies a ticket, extracts data from a PDF, writes a summary. The rest is ordinary rules.
AI in Salesforce
Agentforce, Prompt Builder, Salesforce MCP. A separate site about AI in Salesforce, with a free permissions audit before you switch anything on.
Why me
A developer who connects systems every day
CRM, website, email, databases, APIs. Most of an AI implementation is plumbing your data in. I do that myself, with no third party in between.
Permissions first
AI sees only what the account it runs on sees. I start with a separate account and a minimal scope, not with choosing a model.
A prototype on your real examples
No slides. Within 1–2 weeks you get a working version on your real emails, reports or documents, and you check the results.
One measurable task
Not “AI in the company”, but, say, the report that takes three hours every Monday. After a month you know whether it works.
Tools chosen for the task
Claude, OpenAI, Microsoft 365 Copilot, Gemini, Agentforce. I choose for the task and for what you already pay for, not the other way round.
Care after go-live
Models and tools change. I make sure the report still arrives on Monday and the assistant does not start making things up.
In numbers
- 6 yearson the Salesforce platform, since 2020
- 4Salesforce certifications
- 3 countriesPoland, Germany, Sweden: companies I have worked for
- 5,000+active users of my Android apps
- ~75tools in my own MCP server
- 1–2 weeksto a working prototype
How I work
A call
One task that hurts the most today. How often, who does it, where the data comes from.
A prototype in a week or two
On your real examples. You check the results before anything goes further.
Permissions and data
A separate account for AI, read-only access wherever possible, an activity log.
Go-live and a month of watching
Fixes, measuring the time, deciding what comes next.
Care
Models and tools change. Someone has to keep an eye on it.
What this offer is built on
Three situations from my current work. Company names on request during a call, where clients have agreed.
A software company: AI in finance, accounting, administration and IT
- Situationmy day job as Salesforce System Owner
- Roletook part in rolling out AI across four departments
- LessonAI works where the task is single and measurable
I saw from the inside what works and what does not: which tasks AI handles well from the first week, where a person has to approve, and why permissions have to be tidied before anything is switched on. That is the experience I bring to smaller companies.
My own MCP server and a system written by AI
- MCP serveraround 75 tools, exposed to the Claude assistant over OAuth 2.1
- SystemAI wrote nearly all of the code
- My rolespecification, review, tests
I use every day what I propose to clients: an assistant connected to my own data through MCP, with its own account, scope and log. I know what breaks, because it broke for me first.
Three companies under ongoing care: CRM, website, mobile app
- Trading companySalesforce, ongoing support, about 12 hours a month
- Company with Salesforce and WordPressongoing changes, website connected to the CRM
- Company with a mobile appfixes every two months
These are the systems AI gets connected to: a CRM, a website, an app’s database. I know them from the inside because I look after them week in, week out. An AI implementation starts with someone knowing where the data is and who has access to it.
References available on request, with my clients’ agreement.
A developer by the hour, an AI agency, an off-the-shelf tool, or nothing?
For a small company that wants one task sorted, what matters is time to a first result, access to your data, and who looks after it once it is live.
| Developer by the hour (me) | AI agency | Off-the-shelf SaaS tool | Doing nothing | |
|---|---|---|---|---|
| Starting point | One task and your real data | Workshops and an “AI strategy” | What the tool can do, not what you need | A person keeps doing the task, every week |
| Connecting to your systems | I do it myself: CRM, email, databases, website | Often a separate integrator or subcontractor | Only what it has a ready connector for | — |
| Permissions and data | Separate account, read-only, logged. Set up and documented | Depends on the team | Usually asks for full access to your mailbox or drive | Staff paste data into personal chat accounts |
| Time to first result | A prototype in 1–2 weeks | After an analysis phase | Immediately, if it fits; never, if it does not | — |
| Cost | Fixed price for the prototype, then time worked | A project with a budget running for months | Per-user subscription that grows with the team | Hours every week, invisible on any invoice |
| After go-live | A month of observation, then care by the hour | Support under the contract | The vendor’s help desk | — |
From the knowledge base
Short pieces on implementing AI in a small company without surprises.
Anatomy of the Monday report
The Monday report is the most common first AI project in a small company, because it is measurable: you know how long it took, and you can see when it arrived. Here is what it is made of, step by step.
What not to paste into AI
The biggest AI risk in a small company is not in the model. It is in the employee who, at ten to five, pastes a client contract into a personal chat account to “summarise it quickly”. Here are rules you can introduce in a week.
MCP on a read-only account
Connecting an assistant to company data takes fifteen minutes these days. Doing it so that nobody regrets it six months later takes an hour longer. That hour is a separate account, read-only access and a log.
Who does the work
I work with AI every day: Claude Code as a developer, Claude and OpenAI in production automations, Agentforce and Prompt Builder in Salesforce. I built my own MCP server with around 75 tools, exposed to the Claude assistant over OAuth 2.1, and a system where AI wrote nearly all of the code while I owned the specification, review and tests. At the company where I own Salesforce, I took part in rolling out AI in finance, accounting, administration and IT.
I have been building on Salesforce since 2020. Today I am the sole owner of the Salesforce org at a Swedish software company: development, permissions, integrations, production releases. Before that, two years in a Vaillant Group team (loyalty programme, large Apex API, installer portal) and three years at Britenet for clients in education, insurance and finance. Salesforce certifications: Platform Developer I, Administrator, Platform App Builder, Sales Cloud Consultant.
Who I work with today. A trading company whose Salesforce was implemented by an outside vendor: ongoing support, about 12 hours a month. A company with Salesforce and a WordPress site: 1–2 hours of changes a month and a larger job every six months, including connecting the site to the CRM. A company with its own mobile app: fixes every two months. Names on request during a call.
Two neighbouring topics
AI training for teams
If you would rather your team learnt to work with AI themselves: workshops on your own tasks, with data rules built in.
Free Salesforce audit before AI
On Salesforce? Before you connect AI to it, find out who can do everything in your org. 10 checks, report within a week.
Is there a task someone at your company does by hand every week? Describe it in one sentence.
I reply within one business day. The first call is free of obligation.
Describe what you want automated
One sentence about the task is enough to start. You do not need to know the names of any tools. I reply within one business day.