The AI Guide for Business

Chapter one is built around this question. The short version is to start with the work your team repeats every week and quietly resents: drafting the same kinds of emails and documents, answering the same customer questions, pulling information out of PDFs, or routing requests to the right person. Those are the tasks where AI reliably pays for itself. The guide gives you a checklist for scoring your own processes and explains why the first project should be small, measurable, and low risk rather than a company wide rollout.

The guide gives a range for each type of system rather than one number. Team access to a tool like ChatGPT or Claude runs roughly $20 to $30 per user per month. A focused automation that connects AI to the software you already use is typically $3,000 to $25,000. A RAG system that answers from your own documents is usually $10,000 to $50,000, and a production ready AI agent is $15,000 to $75,000 or more. Those ranges come from quotes and builds we have done for companies across Metro Vancouver, and the guide explains what pushes a project toward the low or high end.

Faster than most people assume for the first project, slower than vendors promise for the ambitious ones. A team rollout of an LLM tool takes one to three weeks including training. A first automation is two to six weeks. A document search system is four to ten weeks, and most of that time goes into cleaning up the source material rather than writing code. Agents are the longest at six to sixteen weeks because they need testing and guardrails before you can trust them with real work. The roadmap chapter lays these out as a sequence so you can plan a year of adoption rather than a single project.

An automation follows a fixed path: a trigger fires, the AI does one defined step, and the result lands somewhere. An agent is given a goal and decides for itself which steps to take, such as looking up an account, checking a policy, drafting a reply, and logging what it did. Most businesses should start with automations because they are cheaper, faster to build, and easier to check. Agents earn their cost when the work varies too much for a fixed script, like the billing audit we built for a property management company in North Vancouver that had to reconcile leases against invoices with no two cases alike.

Yes, and this is the single most requested system we build. It is called retrieval augmented generation, or RAG, and chapter five explains it without the jargon. Instead of relying on what the model learned during training, it first searches your contracts, manuals, past projects, or policies, then writes an answer based only on what it found and shows you the source. Law firms, engineering consultancies, and property managers around Vancouver use this to search years of files in seconds. The guide also covers what makes documents ready for it and why that preparation is most of the project.

It depends on the plan you are on and what you paste in, which is exactly why chapter eight exists. Consumer versions of these tools can use your inputs for training, business tiers generally do not, and neither is appropriate for some categories of client data without extra controls. The guide gives you the handful of questions to answer before any rollout: what data is involved, where it is stored, who can see it, and whether your privacy obligations under BC law allow it to leave the country. For businesses that handle legal, financial, or health records, it also introduces on premise deployment, where the AI runs on hardware in your own building.

Because almost nobody has been shown how to use it well. Chapter two is twelve practical techniques that make an immediate difference: giving the model a role and context, supplying examples of what good looks like, breaking big requests into steps, asking it to critique its own draft, and building reusable prompts for tasks your team does every week. Those same techniques are the core of the training sessions we run for teams in Burnaby, Surrey, and the rest of the Lower Mainland, and the guide is a reasonable way to preview whether your staff would benefit from a proper session.

Some of it you can absolutely do yourselves, and the guide is honest about which parts. Rolling out an LLM tool and training your team on it is well within reach for most businesses. Building an automation is doable if someone on staff is comfortable with tools like Zapier or Make. Where outside help tends to pay for itself is in choosing the right first project, connecting AI to your existing systems safely, and anything involving your own documents or customer data. The roadmap chapter is designed so that you can tell which stage you are at and what you would be hiring for.

The guide focuses on what to build and what it costs, and we keep the funding detail on the blog because programs change. The main one worth knowing about is the BC Employer Training Grant, which can cover a large share of the cost of training staff on AI tools, and there are federal programs that apply to certain types of technology adoption. If you are a small business in the Lower Mainland weighing training against a custom build, it is worth checking what is available before you decide.

It was written for owners, managers, and department leads at companies with roughly ten to two hundred staff, in industries that run on documents, correspondence, and repeat processes. That describes a lot of the Metro Vancouver economy: real estate and property management, law firms, accounting and financial practices, engineering and technical consultancies, construction, and logistics. The examples in the guide are drawn from that kind of work. If you run a very small operation or a highly specialised technical business, you will still find the definitions and cost ranges useful, but the roadmap assumes you have at least a few people whose time you are trying to free up.

It is free with no obligation. The PDF is emailed to you within a few minutes of entering your address. Afterward you may hear from us occasionally with something useful, such as a new article or a case study, and if you want to talk through how any of it applies to your business, the initial consultation is also free. One click unsubscribes you at any time and we do not share your address.