The Setup
Running a solo AI consultancy from London has advantages that I didn’t fully appreciate before starting. London sits in a time zone that overlaps with both European and East Coast US business hours. The city’s tech ecosystem provides a steady stream of networking opportunities. And the concentration of financial services, legal, and consulting firms creates natural demand for AI solutions.
Finding Clients
What Works
- LinkedIn content: Regular posts about AI projects and insights generate the most inbound leads
- Referrals: Happy clients introduce you to their network. This is the highest-conversion channel
- Speaking: Conference talks and workshop delivery build credibility and visibility
- Teaching: University lecturing positions create unexpected business connections
What Doesn’t Work
- Cold outreach: Almost never works for high-value consulting
- Generic AI content: “10 Ways AI Will Transform Your Business” doesn’t attract serious buyers
- Competing on price: Racing to the bottom attracts the wrong clients
Managing Projects
The Solo Advantage
As a solo consultant, I can be deeply embedded in a client’s problem without the overhead of managing a team. Every hour they pay for is an hour of direct expert work. There’s no account manager, no junior developer learning on their dime, no coordination overhead.
The Solo Challenge
The obvious limitation: I can only work with 2-3 clients simultaneously. This means being selective about engagements and charging appropriately for the focused attention each client receives.
Tools of the Trade
My essential stack for running the consultancy:
- Communication: Slack (with clients), email (for formal stuff)
- Project tracking: Linear for task management
- Time tracking: Toggl, with detailed descriptions for every entry
- Development: VS Code + Cursor, Google Cloud, GitHub
- Invoicing: Xero, weekly invoicing cycle
The Economics
Solo consulting is economically attractive if you can maintain consistent utilization. The math is straightforward: charge a rate that reflects your expertise, maintain 60-70% utilization (the rest goes to admin, learning, and business development), and you’ll earn more than most salaried positions while working fewer hours.
The catch: you need to be comfortable with income variability and the responsibility of finding your own work. Not everyone is.
Lessons Learned
- Specialization beats generalization: “AI consultant” is too broad. “RAG systems and legacy code modernization” attracts the right clients
- Say no to bad fits: The most important skill in consulting is turning down work that isn’t right
- Invest in learning: I spend 20% of my time learning new tools, techniques, and domains. This is not overhead – it’s what keeps me valuable
- Document everything: Every project produces reusable knowledge. Capture it systematically