There is a version of this article that lists every AI product that launched this year and gives each one a star rating. This is not that article.
What follows is the shortlist — the tools that are open on my screen on a regular working week, what I actually use them for, and where they fall short. No hype. No affiliate links I can’t stand behind.
The honest context first
I work across programme delivery, PMO governance and technology strategy. My use cases are: structuring complex thinking quickly, processing meeting outputs, drafting stakeholder communications, and researching topics I need to get up to speed on fast. Your mileage will vary based on your role.
ChatGPT Plus
Still my primary thinking tool. I use the GPT-4o model for: structuring reports and governance papers, rewriting dense prose into plain English for executives, sense-checking project risk registers, and generating options when I’m stuck. The o1 reasoning model is genuinely useful for complex dependency analysis — it thinks before it answers.
What it isn’t: a replacement for domain knowledge. You still need to know what good looks like to evaluate what it gives you.
Notion AI
I moved my knowledge base to Notion about two years ago. The AI layer added on top has changed how I use it. Most useful feature: “ask your workspace” — querying notes, documents and meeting records with natural language. For anyone managing a complex portfolio, having a searchable, AI-queryable knowledge base is genuinely transformative.
Otter.ai
I use this for meeting transcription. It connects to Google Meet and Zoom, transcribes in real time, and produces a summary after the call. The summary quality is uneven, but the searchable transcript is the real value — being able to search “what did we decide about the budget approval” three weeks after a meeting is worth the subscription alone.
Perplexity
My replacement for Google for most research tasks. It synthesises sources and gives you a cited answer instead of ten blue links. I use it for: understanding regulatory changes quickly, researching vendors before procurement decisions, and getting up to speed on a new technical domain. The Pro version adds more sources and better models.
What I don’t use (and why)
I have tried: Copilot (too shallow for complex tasks), Gemini (improving, but not there yet for my use cases), Claude for everything (strong at writing and reasoning, but I don’t reach for it as my daily driver — yet). This will change. This space moves fast.
The honest summary
AI tools don’t make you better at your job. They make you faster at the parts you’re already good at — and they expose the parts where you’re unclear in your thinking. If you give a vague brief to an AI, you get a vague output. The discipline of being specific is the actual skill.
Start with one tool. Use it properly for a month before adding another.
UserGenix 
Leave a comment