How Hannah Uses AI to Scale Her AI Consultancy

Frame is an AI and Data Science consultancy that also builds and runs its own AI products. We take a deliberate approach to product development: when we see a consultancy solution delivering repeatable value, we turn it into a scalable product.

That balance between real-world delivery and product innovation is what sets us apart.

This became very clear over the last 18 months. Previously, companies came to us knowing they needed specialist Data Science capability. That mindset has shifted dramatically. Now they are looking to us for reassurance, honesty, and clarity on what actually matters.

We usually recommend starting with a simple AI feasibility study. These conversations take longer now, but that is intentional — we care deeply about the ethical use of AI. It would be disingenuous to propose a six-figure build to a company better served by a lightweight dashboard or automation.

There is a huge amount of scaremongering around AI replacing humans. That is not how we see it. AI empowers people to do more; it makes teams faster, more accurate, and more effective.

However, with that in mind, AI is powerful and gets better minute by minute. We are very conscious of that.

Not every consultancy solution is transferable, but occasionally we deliver something with broader application. Frame Rank came from exactly that place. One of our long-standing partners identified that customers were starting to shop for products inside LLMs.

With OpenAI merchant feeds already live in the US, we could see where the UK market was heading and moved quickly. Rank did not exist before October, and it is now being used by Enterprise clients as well as entry-tier customers. That validation came from rapid iteration, real marketing tests, and real usage measured in weeks rather than months.

It has significantly reduced delivery time. Consultancy proposals used to take weeks. Now AI sharpens language, removes doubt, and works with full context from recorded meetings and scoping sessions. We even feed whiteboard photos into proposals once the solution is sketched.

As long as Liam and I spend a couple of focused hours defining the approach, AI handles the heavy lifting. Every proposal is still reviewed by us, but a 40-page proposal with a detailed appendix no longer takes weeks.

AI is cheap, fast, and increasingly accessible. Non-technical people can build with low-code tools. My degree is in English Language and Literature, not STEM, and yet I am coding daily. As a result, high day rates for technical expertise will face downward pressure. Advisory businesses will need to rethink pricing models to remain competitive.

Six years ago, I would have said technical knowledge as a woman in tech felt like armour. Today I see it differently. Commercial awareness is what actually gives you power. You can be the most technically brilliant person in the room, but if you cannot articulate value, risk, ROI, or why a decision matters to the business, your voice will still get talked over. Technical skills can be learned, outsourced, or automated. Commercial judgment cannot.

Network relentlessly. I joined a free NatWest Accelerator programme early on, and the connections I made there have been pivotal for over five years.

Imposter syndrome is real. Realising others feel it too was incredibly empowering. Ask for help, protect your wellbeing, and listen to your body. It took me four years to learn that rest is not optional. Therapy, meditation, and Kundalini Yoga have been game changers. If you are at your best, your business will be too.

Lack of resilience. AI ideas can sound brilliant in theory and fail in reality. Fail fast, reflect, and try again.

Also, underestimating how many hats you must wear early on. You will not be brilliant at everything, and that is fine. My biggest mistake was doubting myself and letting others talk over me. Tech is still male-dominated, and I held back early on. Liam constantly reminded me that my perspective mattered. He was right.

The first 90 days would not be about building anything shiny. They would be about listening, selling, and proving I can create value before asking for investment or scale. In the first 30 days, I would talk to people constantly, former colleagues, founders, operators, and not pitch a product but ask what is broken and what is quietly costing money.

Days 30 to 60 would be about turning one of those problems into paid work. One client, one use case, one clear outcome. The final 30 days would be about pressure testing what I learned. Only then would I think about hiring, building a product, or raising money. Momentum comes from relevance, not noise.