AI training for Chicago teams

001/ Scope4 parts · 1 rollout

• Role-specific workshopshands-on, on your real tasks
• AI acceptable-use policyapproved tools, data rules, review
• Prompt standardsa shared, tested prompt library
• Copilots & agentsfor workflows worth automating
• Voice agentslike Sofia, our AI receptionist
• Follow-up sessiona few weeks after the workshops

The goal isn’t for your team to “learn AI.” It’s for the sales rep to prep a call in five minutes instead of thirty, for the operations lead to turn a messy voice memo into a clean checklist, and for everyone to know — without asking — what they may and may not paste into a chatbot. Skills and rules arrive together, so adoption doesn’t outrun judgment.

Before a workshop we collect a handful of real tasks from the group (anonymized where needed). In the room, people work on those tasks on their own laptops, with their own accounts, and we show them how to check the output — because the most important AI skill is knowing when the answer is wrong. For a head start, read how to get started with AI in your business.

AI acceptable-use policy and governance

An AI acceptable-use policy tells employees which AI tools are approved, what data must never be pasted in, and when a human has to review the output. We draft one for your business as part of the training, so the rules and the skills arrive together.

• Approved toolsand which accounts to use them with
• Data that never goes incustomer personal data, passwords, confidential client files
• Human reviewwhat must be checked before it’s sent
• Disclosurewhen to tell a customer AI was involved
• Business accounts onlyno personal chatbot logins for work
• Owner & review datewho updates it, and when

The policy is written in plain English, fits on a few pages, and matches the tools you actually use. Want to see the shape of one first? Start with our free AI acceptable use policy template. If AI governance is part of a wider plan — security, vendors, budget — it fits naturally into an IT strategy engagement.

Prompt standards your whole team shares

Prompt standards are a shared library of tested prompts plus one simple structure for writing new ones, so good results don’t depend on who happens to be “good at AI.”

ROLE     You are a sales assistant for [our company].
CONTEXT  Account notes and the last 3 emails are below.
TASK     Summarize where this deal stands and list 3 questions to ask.
FORMAT   5 bullets, then the 3 questions. Plain English.
RULES    Use only the notes below. Say "unknown" if unsure.

Every entry in the library has an owner, a “good output looks like” example, and notes on what to check. It lives wherever your team already works — a shared doc, your intranet, or the saved-prompt feature of the tool you use — and it grows as people find prompts worth sharing. Ready to go further? Our guide to building your first AI workflow shows the next step.

Case study: Sofia

• The problemMissed calls
• Answers24/7
• LanguagesEN + ES
• Price in 312 Scheduling$0.25/min

Barbers can’t pick up the phone mid-cut, so calls go unanswered and chairs stay empty. Sofia answers every call, books and cancels appointments against the shop’s real schedule, and switches between English and Spanish mid-call when the caller does. It’s a small, well-defined job with clear rules — exactly the kind of workflow an agent does well.

Sofia, our AI receptionist →

portal.312itconsulting.com

AI speaking practice in Bosanski

In Bosanski, our Bosnian-learning app, learners talk out loud: Whisper speech recognition listens, and Razgovor, an AI speaking conversation, answers back. It’s scored on being understood, not on being perfect.

Behind the scenes, AI drafts new lessons — but nothing publishes until a teacher approves it. That’s the same “human reviews the output” rule we write into every client’s AI policy, applied to our own product.

Other agents we build: an inbox triage copilot that drafts replies for a human to approve, a document assistant that answers questions from your own policies, and small automations that move data between tools. See real examples in AI agents for small business.

1. Discovery

   Which tools you have, which roles need what, and which data is sensitive. We collect real tasks for the workshops.

2. Policy

   We draft your AI acceptable-use policy with leadership, so the rules are ready before the training starts.

3. Workshops

   2–3 hours per role group, on-site or live online, hands-on with your own tools and tasks.

4. Prompt library

   The best prompts from the workshops become your shared library, with a follow-up session a few weeks later.

How much does AI training cost?

A single on-site workshop for up to about 15 people typically starts from about $2,500. A full rollout is quoted in writing based on team size, number of sessions, and whether you need a policy, a prompt library or a custom agent.

Typical prices, not quotes. Every engagement gets a written scope and quote before work starts. Timelines are estimates and subject to change based on scope.

If you’re choosing between them, we’ll help you pick based on what you already pay for and where your files live — a Microsoft 365 business usually gets more from Copilot, a Google Workspace one from Gemini. Our Microsoft Copilot guide covers the Microsoft side in detail.

Who it’s for

AI training is for small and mid-size businesses whose staff already use AI unofficially, or want to, and whose owners want it used safely and well.

• People using AI on personal accountsbring it into the open, safely
• Leaders who need a policybefore something goes wrong
• Teams with repetitive writingproposals, reports, replies
• Businesses with missed callsa voice agent like Sofia