The Personal AI Checklist: Are You Set Up Right?

10 questions to diagnose your AI setup. If you're missing more than 3, you're leaving serious productivity on the table.

You're using AI. That's good.

But are you using it well? Or are you stuck in the productivity trap, where AI theoretically helps, but in practice you're spending more time wrestling with tools than getting work done?

This checklist will tell you.

Go through these 10 questions. For every "no," you've found a gap. For every gap, there's a fix. And if you're missing more than three, you're ready for a real AI system, not just a collection of tools.

The 10-Point Personal AI Checklist

Does your AI remember you across sessions?

If you're re-explaining who you are, what you do, and how you work every time you open a new conversation, you don't have memory. You have amnesia.

The fix: Build a persistent knowledge base (Notion doc, custom GPT with uploaded files, or vector database) that stores your profile, projects, and preferences. Your AI should pull this context automatically.

Can your AI access your email without copy-paste?

If you're manually pasting emails into ChatGPT for triage or drafting, you're the integration layer, and that's a waste of your time.

The fix: Use Gmail API or Outlook API to let your AI read, draft, and send emails directly. Zapier/Make can handle simple workflows; custom code is better for complex needs.

Does your AI know your calendar and upcoming commitments?

If your AI can't see your schedule, it can't help you prep for meetings, block focus time, or suggest when to tackle tasks.

The fix: Connect Google Calendar or Apple Calendar via API. Your AI should be able to read events, suggest prep work, and (if you want) create/modify events.

Can your AI reference your notes and documents?

Your Notion workspace, Obsidian vault, or Google Drive is full of valuable context. If your AI can't see it, you're constantly re-explaining things that already exist.

The fix: Index your notes/docs in a searchable knowledge base. Custom GPTs support file uploads. Advanced: use vector databases to enable semantic search across all your content.

Do you have automations that run without you?

If you're manually triggering the same AI prompts every day (email triage, meeting prep, research summaries), you're doing work that shouldn't require a human.

The fix: Set up scheduled automations. Use cron jobs, Zapier schedules, or Make scenarios to trigger AI tasks at specific times (e.g., 7 AM email summary, 6 PM next-day prep).

Is your AI tuned to your specific communication style?

Generic AI output feels… generic. If every email draft or doc summary sounds like it came from a corporate template, your AI doesn't know your voice.

The fix: Train your AI on examples of your writing. Upload past emails, reports, or memos. Use feedback loops: when you edit AI drafts, log what you changed and why.

Do you have privacy controls you actually understand?

If you're pasting sensitive client data, legal docs, or proprietary info into public AI tools, and you don't know exactly where that data goes, you have a security problem.

The fix: Use private AI infrastructure: self-hosted models (Ollama, LM Studio), enterprise APIs with no-training guarantees, or fully encrypted pipelines. If you can't explain your data flow, don't trust it.

Can your AI handle multi-step tasks end-to-end?

If you're asking your AI to "research X," then separately asking it to "summarize findings," then manually combining everything, you're doing orchestration work that should be automated.

The fix: Build multi-agent workflows. One agent researches, another synthesizes, another formats. They hand off context automatically. Tools: LangChain, custom scripts, or workflow automation platforms.

Does your AI system work across devices?

If your AI setup only works on your laptop, what happens when you're on your phone, tablet, or a different computer? You lose all the context and continuity.

The fix: Use cloud-synced knowledge bases (Notion, Airtable) and web-accessible AI interfaces (custom web apps, OpenAI API, etc.). Your AI should follow you, not be tied to one device.

Are you confident your AI setup won't break when models update?

AI is evolving fast. If your setup is tightly coupled to one model version or provider, what happens when they deprecate the API or change pricing?

The fix: Build provider-agnostic architecture. Use abstraction layers (LangChain, LiteLLM) so you can swap models without rewriting everything. Have fallback options.

How to Score Your Setup

10/10: You have a real personal AI system. You're in the top 1% of AI users.

7–9/10: You're on the right track. You have good foundations, but there are clear upgrade paths.

4–6/10: You're in power-user territory, but hitting limits. You know what's possible; you just haven't built it yet.

1–3/10: You're using AI tools, not an AI system. You're getting some value, but leaving 80% of the potential on the table.

0/10: Either you just started (great!), or you're using AI in the most manual, lowest-leverage way possible (time to upgrade).

What to Do Next

If you scored 7 or above: You're doing well. Focus on optimizing what you have. Fix the remaining gaps one at a time.

If you scored 4–6: You're ready for the next level. You need architecture (persistent memory, workflow automation, and integration), not just better prompts.

If you scored 0–3: You have two paths:

  1. DIY route: Start with the easiest fixes (custom instructions, uploaded knowledge base, calendar integration). Build up over time.
  2. Hire someone: If your time is worth more than the 40–60 hours it'll take to build this yourself, work with people who've done it before.

Reality check: Most people think they want to build it themselves. Most people hit a wall at API integrations, security architecture, or ongoing maintenance. If you're technical and have the time, go for it. If you're not, or if 60 hours sounds like hell, hire experts.

The Bottom Line

AI isn't just about having access to ChatGPT. It's about building a system that:

This checklist shows you where the gaps are. The fixes are all doable: some in an afternoon, some in a week, some requiring real engineering.

The question isn't whether you can build a proper AI system. It's whether it's worth your time to do it yourself, or whether you'd rather have someone hand you the finished product and get back to your actual work.

Ready to Put This Into Action?

Book a free 15-minute AI audit. We'll review your setup against this checklist, identify your biggest gaps, and give you a personalized action plan, no strings attached.

Book Your Free AI Audit

- The Catalyst Team