# AI strategy and consulting for Alaska organizations K.AI Consulting is an AI consulting practice in Anchorage, Alaska, run by Kai Myers. I help Alaska businesses, agencies and small teams decide where AI is worth using, choose the tools, train the people who will use them, and build the parts that can't be bought. Most engagements start with one workflow and a 30-minute call. Published 2026-10-01 · Updated 2026-10-01 · Author: Kai Myers, K.AI Consulting, Anchorage, Alaska Canonical page: https://kaiconsulting.ai/ai-consulting ## What you can hire me for ### AI strategy and advising (Scoped and priced on the first call) For an owner or manager who keeps hearing the team should be using AI and wants a plain answer before spending anything. I sit down with the people who do the work, go through what a normal week involves, and sort it: jobs a tool you already pay for can do, jobs worth buying something for, jobs worth building, and jobs to leave alone. - A working session with the people who do the work - A short written plan: what to do first and what it costs to keep running - Which steps keep a person checking the output - A second opinion on a vendor quote or product, if you have one ### AI primer (60 to 90 minutes · any group size) A clear introduction for a whole team or a board: what AI is and isn't, which tools are mature, what is realistic in your line of work, and a few things people can use the same day. Live demonstrations on real examples. - What AI is (and isn't), without the hype - Which AI tools are mature and which are still experimental - Realistic expectations for AI in your industry - Quick wins you can implement immediately ### Hands-on AI workshop (Half day or full day · 1 to 10 people) Your team, your documents, current tools. We assess what you need beforehand, set the tools up together, and build the prompts, templates and guardrails for the tasks you actually do. The technical half covers connecting AI to your own data sources. - Extended needs assessment - Multiple tool implementations - Team training and documentation - 60-day follow-up support - Custom prompts and templates ### AI built into a tool (Fixed-fee sprint or production build) When the right answer is software, I build it: an assistant that answers from your own documents, a drafting step inside an existing workflow, or a tool with no AI in it at all because fixed rules do the job better. See the engagement options on the Services page. - Defined scope and fixed source material for anything a model does - Output a person can check - Ordinary, testable software around it - Documentation and a recorded handoff ## Where AI went in three real jobs ### No AI in the product: lab data review A remediation contractor's weekly lab review is checked against cleanup levels and a state checklist. That has to give the same answer every time, so the tool is plain rules encoded from the firm's own cheat sheet. AI was used heavily to build it and not at all when it runs. https://kaiconsulting.ai/work/lab-data-review-tool ### AI with fixed source material: a model you can question The Dixon Glacier workbench includes an assistant that answers questions against the model's own documentation, so a reader can interrogate the work directly. It also publishes a disclosure of where AI assistance was and was not trusted in building the model. https://kaiconsulting.ai/work/dixon-glacier ### AI doing one narrow job: plain-language regulations and forecast summaries A regulation simplifier and an avalanche forecast summarizer both use a model at runtime. Each runs on a system prompt, a skills file and reference documents, so it does the one job it was given with the source material it was told to use. https://kaiconsulting.ai/services ## A sample AI primer (60 to 90 minutes) - 0:00 - 0:10 Welcome & Who I Am: My background, how I came to AI, and a roadmap for our session - 0:10 - 0:25 AI Demystified: What AI is, what it isn't, how it works under the hood, and common misconceptions - 0:25 - 0:40 Types of Models: Fast models vs. thinking models, context windows, and the basics of prompt engineering - 0:40 - 0:55 Simple Tools for Any Industry: Practical wins anyone can use, like integrating Granola for meetings and Comet Browser for smarter searching - 0:55 - 1:10 AI Specific to Your Industry: Exploring specialized tools or discussing how to build custom solutions for your unique use case - 1:10 - 1:30 Q&A & Discussion: Open floor for questions and conversation ## A sample full-day AI workshop - 8:00 - 8:15 Welcome & Coffee (Whole Team): Introductions, goals for the day, and workshop overview - 8:15 - 9:00 AI Fundamentals (Whole Team): Types of models (fast vs. thinking), model selection, prompt engineering, context engineering, and understanding where AI excels vs. where it falls short - 9:00 - 9:30 Needs Assessment Review (Whole Team): Review pre-workshop survey results and prioritize focus areas for your organization - 9:30 - 9:45 Break: 15 minutes - 9:45 - 10:30 Hands-On with Claude: Prompts & Context (Whole Team): Practice prompting in Claude with real tasks from your work, then load documentation into a Claude Project to see how context dramatically improves results - 10:30 - 11:30 Tool Tour: Kai's Favorites (Whole Team): Hands-on demos of Granola (meeting notes), Comet Browser (AI-powered search), and NanoBanana (image editing). Set up each tool and explore which ones resonate - 11:30 - 11:45 Break: 15 minutes - 11:45 - 12:45 Working Lunch (Whole Team): Continued exploration with informal Q&A - 12:45 - 1:45 Getting Your Data In: MCP, CLI, API (Technical Team): Connect Claude to your real data sources: Model Context Protocol servers, command-line workflows, and direct API integrations. Pick the right tool for each job. - 1:45 - 2:30 Apply: Field Data & Compliance Automation (Technical Team): Set up Claude Code in VS Code, then build prompts for Phase I/II report drafting, ADEC regulatory lookups, and field sampling data QA/QC - 2:30 - 2:45 Break: 15 minutes - 2:45 - 3:45 Creating and Using Skills (Technical Team): Build reusable Claude Code skills for your team's recurring workflows, including how to author them, when to invoke them, and how to share them across projects - 3:45 - 4:15 Apply: Construction & Project Documentation (Technical Team): Use Claude Code to generate daily construction logs, draft RFIs and change order responses, and automate equipment/materials tracking summaries - 4:15 - 4:30 Wrap-Up & Next Steps (Technical Team): Documentation handoff, troubleshooting tips, and 60-day follow-up plan ## Questions people ask first ### We want to use AI but don't know where to start. What do we do first? Pick one task somebody repeats every week and would happily never do again. Bring it to a 30-minute call. Often the first step is a tool you already pay for, used properly. Sometimes it's a small build. Sometimes the answer is to leave it alone, and I'll say so. ### Should we buy an AI product or build something? Buy when a product already does the job and you can live with how it works. Build when the rules are your own, the data can't leave the building, or the work happens where there is no signal. I'll tell you which I think it is before you spend anything. ### Can AI be trusted with regulatory or compliance work? Not on its own. For anything that gets submitted or audited I build fixed rules that give the same answer every time, and keep AI to development and to drafts a qualified person reviews. The tools are decision support. The professional still signs. ### Can we have an assistant that answers from our own documents? Yes. The work is in choosing the source material and holding the assistant to it. The Dixon Glacier workbench has one that answers against the model's own documentation. ### Do you only work with environmental and field organizations? No. That's the work I came from, so I know labs, remediation contractors, agencies, research groups and avalanche centers best. The same approach fits any office or crew that runs on forms, spreadsheets and reports. ### Who is this for? Small and mid-sized Alaska organizations without a developer or a data team on staff. ### What does it cost? Everything is quoted up front after the first call, so you know the number before any work starts. Builds are fixed-fee or fixed scope with milestone payments. ### Where do you work? Anchorage, Alaska, and remotely across the state. On site for workshops and kickoffs when it helps. Contact: kai@kaiconsulting.ai · 916-955-8064