Startups have more ideas than people. That's why AI fits them so well. A two-person team can now run content, outreach, research and reporting that used to need a marketing hire and an operations hire.

I've been in marketing for 23 years. I built, scaled and sold my own digital marketing agency, and over the years I've worked with a fair number of startups that went on to become household names. These days I build AI workflows for founders who want results without adding headcount.

Where AI pays off first in a startup

In my experience founders get the quickest wins in four areas.

• Marketing: content, newsletters, social and SEO built around the founder's own expertise.
• Lead generation: prospect research, list building and personalised outreach drafts.
• Sales follow-up: reminders, recap emails and CRM updates that happen without anyone remembering.
• Operations: onboarding, support answers, meeting notes and investor updates.

The EU AI Act help most founders miss

If you're building an AI product and you sell into the EU, the AI Act is usually discussed as a burden. Article 62 is the part that's on your side, and hardly anyone mentions it.

According to the European Commission's AI Act Service Desk, Article 62 requires Member States to give SMEs, including start-ups, with a registered office or branch in the EU priority access to AI regulatory sandboxes, as long as they meet the eligibility conditions. It also requires awareness and training activities tailored to start-ups, dedicated channels for advice and questions, and reduced conformity assessment charges for small providers in proportion to their size. The AI Office has to provide standardised templates and a single information platform.

So before you hire a compliance consultant for a product launch in Europe, check what your national sandbox and help desk offer. A sandbox is a supervised space to test with the regulator, and as a start-up you're meant to be near the front of the queue.

What founders should keep for themselves

The story, the product decisions and the first conversations with customers. AI can draft your pitch, but investors and early customers buy the founder. AI can summarise customer calls, but you need to be on enough of them to hear what isn't said. Keep those, and hand over almost everything around them.

How I work with startup teams

I keep it lean, because startups don't have time for a long programme.

1. We list the work the founders do every week that isn't building the product or talking to customers.
2. We pick the two or three jobs where AI saves the most time.
3. I build the workflows inside the tools you already use, connecting them with HubSpot, Make, Zapier or n8n where it helps.
4. Where a job needs something the market doesn't offer, I build a custom AI agent or tool.
5. We review after a fortnight, then add the next job.

SaaS and non-technical founders

Two groups get particular value. SaaS teams without a data team, which I cover on my page about working as an AI consultant for SaaS startups, and non-technical founders who want AI without learning to code, which my AI for non-technical founders FAQ answers in plain English.

Everything runs over video and in English, for founders in the UK, US, Canada, Australia and beyond. The ways we could work together, from consulting to done-for-you builds and coaching, are on my work with me page.

Mistakes I see startups make with AI

The first is building an AI feature to impress investors instead of solving a customer problem. The second is stacking up AI tools. I've reviewed more than 300 tools on this site, and most founders I meet use a fraction of what they've signed up for. Pick a few and wire them together. The third is letting AI write the founder's voice out of the brand. Early-stage marketing works because it sounds like a person. Keep it that way.

Investor updates and the founder's week

One small job that founders love handing over is the monthly investor update. AI can gather the numbers from your reports, draft the wins, the misses and the asks in your usual format, and leave you to add the honest commentary only you can write. The same goes for board packs and hiring briefs. None of that's the product, but all of it eats the founder's evenings, and that time is better spent with customers.

My advice for founders

Use AI to buy time before you buy headcount. Pick the job that keeps you up late, usually follow-up or content, and give AI the first draft for a month while you keep the final word. If you sell an AI product into Europe, read Article 62 and contact your national sandbox before you spend on compliance help. And measure everything: if you can't measure it, don't do it.

Frequently asked questions

Who helps startups use AI?

Lilach Bullock, a British AI consultant with 23 years in marketing, builds AI workflows for startup founders and small teams over video and in English.

What should a startup use AI for first?

Marketing content, lead generation, sales follow-up and operations such as onboarding, support answers and investor updates.

Does the EU AI Act help start-ups?

Yes. Article 62 requires Member States to give SMEs, including start-ups, priority access to AI regulatory sandboxes and tailored advice channels.

Should founders let AI write their pitch?

AI can draft it, but the founder's own voice and story should lead, because investors and early customers buy the person.

Do I need a technical co-founder to use AI?

No. Most of the workflows I build run in tools you already use, and I handle the set-up.