If you're a solo consultant and you're still spending 90 minutes writing a client proposal from scratch, that's a solvable problem. If you're summarizing research into a deck by hand, that's a solvable problem. If your post-project documentation is always the last thing that gets done because it takes forever, that's a solvable problem.
Not all of your problems. But those specific ones, yes.
I want to be upfront about something: this site is nominally about coaching, but I hear from a lot of solo consultants, boutique agency operators, and independent advisors who found their way here because the voice resonated. There's a real overlap between coaching and consulting, and many executive coaches have consulting threads running through their practice anyway. So this one's for the broader category. If you're a coach specifically and want a more tailored take, the best coaching software piece is probably more useful.
For everyone else: here's what I'd actually set up. Not 40 tools. Four or five, each with a specific job.
---
The question you have to ask first
Before you install anything, ask yourself: what's the recurring task in my practice that takes three times longer than it should?
Not "where could AI help generally." The specific, recurring thing that makes you groan when it appears on your to-do list. The one you keep pushing to tomorrow.
For me it was session notes. For most consultants I talk to, it's proposals, or research synthesis, or the post-project debrief document that never gets written because the next project starts before the last one is fully closed. For some it's just the steady drip of email.
The answer to that question determines which tool matters most for you, and which ones you can safely ignore for now.
---
Claude, for everything document-heavy
The honest answer on Claude versus ChatGPT: they're close enough that the debate is mostly noise, except for one thing. If your work involves long documents, Claude handles them better. A 40-page client report, a detailed project brief, a dense RFP you need to extract requirements from. Claude can hold more of a document in its working context without losing the thread. That's the real difference.
I have a setup I call Margaret, built out of Claude, tuned to my practice's voice, my frameworks, my way of structuring client communications. I reference it occasionally. For consultants, the equivalent would be a Claude setup trained on your proposal format, your client language, your typical deliverables.
That setup takes a few hours to build properly. Then it saves you the "writing from scratch" tax on every engagement.
Cost: Claude Pro runs $20/month. If you're doing any significant volume of document work, that math closes fast.
One thing I'd tell you: don't use it to generate generic professional output. Use it to generate output that sounds like you. There's a learning curve there, and it's worth paying attention to. The goal isn't to have AI write your proposals. It's to have AI write your proposals the way you'd write them if you had an extra three hours.
---
Fathom, for meetings you actually need to remember
If you run client calls, discovery calls, project check-ins, the tool that pays for itself fastest isn't the writing tool. It's Fathom.
Fathom records and transcribes your meetings and gives you an AI-generated summary afterward, organized by topic. It integrates with Zoom, Google Meet, and Teams.
Here's why it matters specifically for consultants: client calls are full of information you don't want to lose. Requirements, priorities, things the client said offhand that turn out to be critical three weeks later. Most of us are trying to listen well and take notes simultaneously, which means we're doing both badly.
Fathom lets you be in the conversation. The notes take care of themselves.
I started using a similar approach for coaching sessions (though the ethics there require some care and explicit client consent). For consulting contexts, it's more straightforward. You're not in a therapeutic relationship. You're in a professional services one. Record the call, let the tool capture the key points, spend your energy actually thinking.
Cost: free tier covers a lot. The paid plan is $19/month and unlocks unlimited recordings and team features. If you're solo, the free tier might be enough.
The ROI calculation is simple: how much time do you spend after calls rewriting your notes into something usable? How much do you charge per hour? Do the math.
---
Perplexity, for research that doesn't make things up
Consultants do a lot of research. Market sizing, industry context, competitive landscape, background on a prospective client before a pitch call.
The problem with using ChatGPT or Claude for research is that they'll write you a confident-sounding paragraph that includes a statistic nobody can actually verify, or a citation that doesn't exist. This is a real problem when your credibility is on the line.
Perplexity cites its sources. Not perfectly, not always from the best sources, but it points at actual web pages and lets you check. For preliminary research and background, it's significantly more trustworthy than asking an AI to synthesize from its training data.
I use it the way I'd use a good research assistant: get the lay of the land, identify the real sources, then go verify the things that matter.
Cost: $20/month for Pro, which gets you better models and more sources. The free tier is usable for lighter research. You can probably start free and upgrade when you hit the limits.
One honest caveat: Perplexity is not a substitute for actual research on high-stakes questions. It's a starting point. If you're advising a client on a major strategic decision, you still need to read the primary sources.
---
Notion AI, if you're already in Notion
This one requires a qualifying statement: Notion AI is a useful add-on if you already live in Notion. It is not worth switching your systems to get.
If you're already using Notion for project management, client notes, knowledge management, the AI layer is genuinely helpful. It can summarize pages, draft documents in your workspace, help you find things across your notes. At $10/month added to your Notion plan, it's easy to justify.
If you're not already in Notion, don't start there just to get the AI feature. Use Claude directly. The output will be similar, and you won't have to rebuild your systems.
I mention this only because Notion AI comes up constantly as a recommendation in these roundups, and the implicit assumption is always "and you should move everything to Notion." That's a bigger decision than it sounds. The AI is good. The switching cost is real.
---
The tool most consultants are already paying for
Here's the one that surprises people: AI in Gmail and Outlook has gotten genuinely good over the past year, and most people who already pay for Google Workspace or Microsoft 365 aren't using it.
Google's AI features in Gmail can draft replies in your tone, summarize long email threads, and pull action items from a chain of messages. Microsoft Copilot in Outlook does similar things, plus it integrates with Teams and Calendar in ways that start to feel like actual leverage once you're used to it.
This isn't the flashiest recommendation. But if you're already paying for these tools, the AI layer is either included or available for a small add-on. Before you buy a new tool, check what you already have access to.
The honest version: these features are best for email management and meeting prep. They're not going to replace a well-configured Claude setup for heavy document work. But for the daily overhead, they're worth turning on.
---
What to ignore
Jasper, Copy.ai, and the category of "AI content tools" in general.
These products were built for marketing teams that need to produce a lot of generic content quickly. That's a real use case. It's not yours.
As a solo consultant, you're not trying to produce volume. You're trying to produce work that sounds like you and reflects your specific thinking. The generic content tools optimize for the opposite of that. They make it easy to produce something that sounds like every other consultant, which is exactly the thing you've spent years trying to not sound like.
The consultants I know who've tried Jasper or Copy.ai for client deliverables have all quietly stopped. The output requires so much editing that you'd have been faster writing from scratch.
Save the money. Put it toward Claude and Fathom.
---
A note on scale
Everything I've described here is calibrated for a solo practice. If you're running a 10-person firm, the calculus changes: you need shared systems, version control, more structured handoffs between team members.
But for solos, the overhead of building elaborate shared systems is usually a distraction from the actual work. Keep it simple. Four tools with clear jobs. One month of consistent use before adding anything new.
The consultants who get the most out of AI aren't the ones with the most tools. They're the ones who have two or three things dialed in well enough that they run on muscle memory.
---
The stack, plainly stated
- Claude ($20/month): primary AI for documents, proposals, synthesis, anything that involves a lot of text
- Fathom (free to $19/month): meeting capture and summaries, so you can stop taking notes while trying to listen
- Perplexity ($0 to $20/month): research with citations you can actually check
- Notion AI ($10/month add-on, only if you're already in Notion): in-workspace AI for knowledge management
- Gmail/Outlook AI (check what you're already paying for): email management and thread summaries
If I were starting over today, I'd set up Claude and Fathom in week one and use everything else for a month before deciding whether it was worth the friction. Those two change the economics. The rest are multipliers on top.
The AI writing workflow piece goes deeper on how to actually structure your Claude setup if you want to get into the specifics of that.
What's the task that takes three times longer than it should? Start there.