AI Tools I Use Every Day as a Consultant (and the Ones I Dropped)

AI tool reviews are mostly marketing. Here's the honest version: what I actually use. After 10+ years consulting and testing dozens of tools, I know what sticks and what's hype. I'll show you both.

My workflow has evolved significantly. Five years ago, I was doing repetitive tasks manually. Three years ago, I was trying to use three different AI tools for different tasks. Today, I have consolidated to a focused set that genuinely speeds up my work and saves me thinking time. The tools I am about to describe are ones I use nearly every day on real client projects: not hobby projects or experiments.

Claude Code: My Primary Development and Automation Tool

Claude Code is where I spend most of my time these days when I am building anything. Whether it is rapid prototyping, debugging production issues, writing automation scripts, or exploring a new codebase, Claude Code ability to actually read code, understand context, and generate working solutions has transformed how I work.

What makes it different from other coding assistants: it actually understands architectural context. I can paste a complex system, ask it to find performance bottlenecks, and get back specific, accurate answers rather than generic suggestions. For client work, this translates to maybe 4-5 hours per week of saved development time.

Practical tip: When using Claude Code with a large codebase, start by asking it to draw a dependency diagram of the modules. This 5-minute conversation saves me hours later because I understand the architecture before diving into edits. The tool shows me relationships I would have missed reading the code myself.

Perplexity for Research and Competitive Intelligence

Research used to mean spending time jumping between browser tabs, finding sources, cross-referencing, and synthesizing. Now I use Perplexity for almost all research because it combines AI reasoning with current web data and gives me citations.

For competitive intelligence, this is invaluable. I can ask it to find recent funding announcements for our competitors, identify new product releases, or understand market positioning. It is faster than manual research and I get source material I can actually verify.

Practical tip: Always check the citations. Not because Perplexity is unreliable, but because sometimes the sources it finds are outdated or tangentially related. A 30-second review of the sources saves you from citing something wrong to a client.

Midjourney for Presentation and Marketing Visuals

Every consulting project involves presentations, and presentations need visuals. Midjourney lets me generate custom images for decks without hiring a designer or spending hours hunting for stock photos that are close enough.

I use it primarily for: conceptual diagrams (like system architectures), hero images for presentations, and marketing content. The quality is good enough that it does not scream AI generated, and it is fast enough that if one image does not quite work, I iterate in minutes.

This is probably worth 2-3 hours per month of design work that I would otherwise be doing or outsourcing.

Practical tip: Be specific about style and context. Instead of AI image, write architectural diagram in the style of a technical system design, with boxes and arrows, professional, minimal color palette. The more specific you are, the faster you will get usable output on the first or second try.

Otter.ai for Meeting Summaries and Action Items

Client calls, stakeholder meetings, team syncs. I record them and use Otter to transcribe and summarize. I am not manually taking notes anymore, which means I am actually present in the conversation instead of fighting to keep up.

The summarization feature pulls out key decisions, action items, and next steps. Then I spend 2 minutes reviewing and editing before sending it back to the team. This level of structure around meetings prevents so many I thought we decided X arguments later.

Practical tip: Review the summary before sharing it with others. Otter does a great job, but it will occasionally misunderstand a domain-specific term or miss nuance. A quick read-through ensures the summary matches what was actually decided.

ChatGPT for Document Drafting and Analysis

I still use ChatGPT (not Claude, primarily because I use Claude Code for development work and prefer to keep my workflows separated) for some types of writing. Specifically: drafting client emails, analyzing documents, creating outlines, and working through complex writing where I need to workshop ideas.

For something like analyze this RFP and identify the top 5 risks, I can paste it in, get back solid analysis in 30 seconds, and then spend my time refining the analysis rather than generating it from scratch.

Practical tip: Use ChatGPT custom instructions to set your tone and style. I have it set to sound like me (direct, practical, no unnecessary jargon). This means the output requires less editing before it is ready to share.

Notion AI for Meeting Agendas and Document Organization

I use Notion for almost all my note-taking and project tracking, so it made sense to use Notion AI for this work too. The integration is smooth. I can type rough notes and ask Notion to organize them into agendas, or paste meeting notes and ask it to extract action items.

It is not the most powerful AI tool, but the tight integration with my actual workspace means it is the tool I reach for most when I am already working in Notion.

Practical tip: Use it as a first draft tool. Have Notion AI generate the structure or the first draft, then edit from there. This still saves 20-30% of the work compared to doing it from scratch, and you maintain control over the final output.

The Tools I Dropped (and Why)

I want to be equally transparent about the tools I tried and stopped using. These are not bad tools: they just did not fit my workflow.

Jasper for Marketing Copy

I tried Jasper for client marketing copy when it first launched. The marketing was impressive, but the tool itself required me to feed it into my workflow in a way that was not natural. I would write in Jasper, then export to Notion, then edit in a separate tool. After a few months, I realized I was spending more time managing the tool than it was saving me. ChatGPT in a browser tab turned out to be faster. I dropped it after three months.

Copy.ai

Similar story to Jasper. Copy.ai is positioned as a marketing copy specialist, but in practice, for me, it was doing the same thing as ChatGPT with more friction. The differentiator just was not significant enough to justify another subscription.

Zapier Native AI Features

I tried using Zapier built-in AI to generate certain automation logic, thinking it would streamline my workflow. In practice, the AI suggestions were generic and I ended up writing the logic myself anyway. The promise was good, but the execution was not powerful enough for the types of problems I solve. Now I use Claude Code to write Zapier workflows, which is faster and more reliable.

The Tools I am Watching (But Not Using Yet)

There are a few tools that are interesting but I have not integrated into my daily workflow because they do not solve an acute problem I have:

• GitHub Copilot: Good tool, but since I use Claude Code for development work, adding another AI assistant at the IDE level just adds noise. I might revisit this if I am working primarily in a language where Claude Code is weaker.
• Gamma: AI presentation builder. It is impressive, but for me, the manual control of presentation design is worth keeping. Your mileage may vary here.
• Custom GPTs: These are interesting for specific domains, but they are not powerful enough yet for the custom workflows I have tried to build with them. I will watch this space.

The Real Lesson: Consolidate, Do not Accumulate

The biggest mistake I see people make with AI tools is tool accumulation. They have ChatGPT, and Claude, and Perplexity, and Notion AI, and Copy.ai, and three other tools they signed up for. Then they spend mental energy deciding which tool to use for each task.

My approach is the opposite: find the best tool for your most common task and make it work. Only add another tool if it saves you more time than it costs in context switching. I would rather be excellent at using five tools than mediocre at using fifteen.

The tools I listed are not the best tools: they are the best tools for my specific workflow, client work, and how my brain works. Your stack might look different. The principle remains: choose consciously, use deliberately, and drop ruthlessly when something stops earning its place.

If you are building a consulting practice or trying to scale your own work and want to talk through your AI tool strategy, I am happy to discuss what is working and what is not in your specific situation.

Richard Batt has delivered 120+ AI and automation projects across 15+ industries. He helps businesses deploy AI that actually works, with battle-tested tools, templates, and implementation roadmaps. Featured in InfoWorld and WSJ.

Frequently Asked Questions

How long does it take to implement AI automation in a small business?

Most single-process automations take 1-5 days to implement and start delivering ROI within 30-90 days. Complex multi-system integrations take 2-8 weeks. The key is starting with one well-defined process, proving the value, then expanding.

Do I need technical skills to automate business processes?

Not for most automations. Tools like Zapier, Make.com, and N8N use visual builders that require no coding. About 80% of small business automation can be done without a developer. For the remaining 20%, you need someone comfortable with APIs and basic scripting.

Where should a business start with AI implementation?

Start with a process audit. Identify tasks that are high-volume, rule-based, and time-consuming. The best first automation is one that saves measurable time within 30 days. Across 120+ projects, the highest-ROI starting points are usually customer onboarding, invoice processing, and report generation.

How do I calculate ROI on an AI investment?

Measure the hours spent on the process before automation, multiply by fully loaded hourly cost, then subtract the tool cost. Most small business automations cost £50-500/month and save 5-20 hours per week. That typically means 300-1000% ROI in year one.

Which AI tools are best for business use in 2026?

It depends on the use case. For content and communication, Claude and ChatGPT lead. For data analysis, Gemini and GPT work well with spreadsheets. For automation, Zapier, Make.com, and N8N connect AI to your existing tools. The best tool is the one your team will actually use and maintain.

Put This Into Practice

I use versions of these approaches with my clients every week. The full templates, prompts, and implementation guides, covering the edge cases and variations you will hit in practice, are available inside the AI Ops Vault. It is your AI department for $97/month.

Want a personalised implementation plan first? Book your AI Roadmap session and I will map the fastest path from where you are now to working AI automation.