Practical guidance for teams looking to use AI effectively. I focus on implementation that works, not hype or vendor marketing.
You can't open a news source without someone declaring AI either the future of everything or the downfall of humanity. The media has made a complete mess of understanding what AI actually is and what it can practically do for teams.
Meanwhile, actual teams are trying to figure out how to use these tools without wasting money on expensive subscriptions, compromising their data security, or getting caught in vendor lock-in.
Clear understanding
What these tools are, what they aren't, and where they fit in your workflow
Practical implementation
How to set up models that work for your specific needs and constraints
Cost-effective approaches
Options that don't require enterprise licensing or expensive cloud credits
Data sovereignty awareness
Understanding when local models make sense versus when cloud options are appropriate
I don't push the latest marketing hype. I focus on models that are practical, economical, and can be deployed in ways that make sense for your team.
Run models on your own infrastructure. This approach gives you complete control over your data and eliminates ongoing API costs.
When appropriate, cloud options provide access to more powerful models without the need for specialized hardware investment.
The right choice depends on your specific requirements, regulatory environment, and budget. I help you understand both options and make an informed decision that aligns with your constraints.
LM Studio has become my primary tool for working with local models. It provides a straightforward interface for downloading, managing, and running a wide variety of open-source models without requiring extensive ML expertise.
More importantly, I'm an advocate for the Model Context Protocol (MCP) - a standard that enables AI assistants to connect with external tools, APIs, and data sources in a structured way. This moves beyond simple chat interfaces and enables actual workflow automation.
I've built and open-sourced an MCP Document Processor that helps AI agents work with document formats more effectively. It extracts text, recognizes structure, pulls metadata, and can create documents with styling options.
View the source codePractical implementation
I use these tools daily. I'm not reciting documentation - I'm showing you what actually works.
Custom integrations
Understanding MCP means I can help you connect AI to your existing tools and workflows.
Local-first approach
I can show you how to achieve significant value without sending your data to third-party services.
Real experience
I've built tools, debugged issues, and found workarounds. That's different from reading a blog post.
There's a narrative that AI will replace developers or make teams obsolete. I take a different view. AI is a tool that, when used properly, augments human capability.
Use AI to clarify your thoughts, structure arguments, and communicate more effectively. Your expertise remains - the tool helps you express it better.
Focus AI on specific outcomes. Define the problem clearly, use the tool to explore solutions, and apply your judgment to select the best path.
AI generates options. You decide. Understanding the tool's limitations keeps you in control and prevents blind reliance on outputs.
When I use AI, I'm not having it write my code or compose my emails. That's lazy and produces mediocre results. Instead, I use it to:
The output always passes through my judgment. I understand what it's doing, why it's doing it, and whether it makes sense for my specific context.
Are skeptical but curious
You've heard the hype and you're not buying it, but you suspect there might be something useful underneath.
Value data privacy
You work in industries or environments where sending proprietary data to third-party services is not acceptable.
Need practical outcomes
You're not interested in theoretical discussions. You want to know how to actually use these tools to get work done.
Have constrained budgets
Enterprise licensing fees and expensive cloud credits aren't in your budget. You need options that make financial sense.
Want to build capability
You're not looking for a one-off fix. You want to understand how to integrate AI into your team's workflow sustainably.
That your team will become AI experts
I provide the brain dump and guidance. Whether your team runs with it is up to them.
Instant productivity gains
Learning curves exist. Adoption takes time. I don't sell magic pixie dust.
One-size-fits-all solutions
Every team is different. What works for one might not work for another. I help you find what works for you.
That AI will solve all your problems
AI is a tool. It can help with some things, make others worse, and has no bearing on most issues.
An effective brain dump of what I've learned through practical implementation. Clear explanations of options, honest assessment of trade-offs, and guidance that's grounded in real experience rather than marketing materials.
Whether that's valuable to you is your call. I don't push sales - I have enough work from referrals and ongoing engagements. I do this because I believe teams need better guidance than what the industry currently provides.
Every engagement is different, but here's what typically gets addressed as part of the advisory process.
Let's have a conversation. I don't have fixed packages because every situation is different.
One-Day Intensive
Deep dive into your situation, hands-on setup of local models, practical demonstrations, and documentation of next steps. Good for teams ready to move quickly.
Multi-Week Engagement
Initial assessment followed by weekly check-ins. Implementation support, troubleshooting, and iterative refinement. Good for teams that want sustained guidance through the adoption process.
Monthly Advisory
Ongoing strategic guidance for teams that are already using AI but want to optimize their approach. Focus on advanced patterns and emerging opportunities.
I'm not trying to upsell you or create dependency. My goal is to get you to a point where you're self-sufficient. If you need me again later, great. If not, that's also a successful outcome.
The best way to figure out what makes sense is to have a conversation. No pressure, no sales tactics. Just a discussion about your situation and whether I can help.
Whether it's migrations, Forge app development, or practical AI guidance, I'm here to help. The best way to start is by asking in community.
Or check out my open-source MCP Document Processor to see what I've built.