Experts Share: What To Consider When Running An AI Startup – TechRound

Microsoft and Nvidia have launched a new programme to support AI startups in the UK. They will be helping companies that already have early funding and are looking to further develop their AI products. Startups joining the programme will get practical help with technology, marketing strategies, and connections to investors.

The programme backs startups that want to create solutions to improve everyday life, generate jobs, and contribute to economic growth. Darren Hardman, head of Microsoft UK said, “By supporting local AI start-ups and providing them with the resources they need to succeed, Microsoft and our partners are helping to build a vibrant and sustainable AI community in the UK and create the tech powerhouses of tomorrow.

“These fast-growing companies will create jobs, boost UK economic growth, and create products that make a positive impact on people’s lives. I’m excited to see what innovations emerge from this unique initiative.”

What Kind Of Support Will Startups Get?

Each startup will get dedicated help from Microsoft’s AI Black Belt team, specialists in artificial intelligence and machine learning. Nvidia will also support startups through training sessions at its Deep Learning Institute, alongside access to its computing tools, designed to improve AI project development and performance.

Besides technical support, startups will receive advice on refining their marketing plans and creating opportunities to meet potential investors. GitHub workshops will offer additional resources to further develop their products and skills.

The programme is designed to work for startups across the UK with both online and in-person components, making sure there’s access to companies from different regions.

Who Is Eligible For The Programme?

All AI startups in the UK that have already launched products and are ready for the next phase of growth can join. Microsoft wants to work with businesses that use AI to look at real world problems in industries like finance, healthcare, and environmental technology.

This programme helps companies with early stage funding and gives them the resources they need to progress further. Applications are open from 4 November to 22 November 2024. A list of successful applicants will be announced in early December. The programme itself starts on 22 January 2025 and ends on 5 March 2025.

What To Think About When Running An AI Startup

Of course, funding and support are not the only things to keep in mind when running an AI startup. Experts have taken the time to share everything else that needs to be considered, for the success of an AI startup. Here’s what they’ve shared:

Our Experts:

• Patrick Murphy, Founder & CEO, Togal.AI
• Kamen Brestnichki, Co-Founder, Martian Lawyers Club
• Dima Eremin, CEO, BluedotHQ
• Jim Boswell, CEO, OnPoint Healthcare Partners
• Catalin Zorzini, CEO, TutorAI.me
• Swapnil Shnide, CEO, Zeni
• Erik Schwartz, Chief AI Officer, Tricon Infotech

Patrick Murphy, Founder & CEO, Togal.AI

-AI investment is surrounded by hype, but the real key to success is identifying where AI can deliver the most value for businesses—and focusing on those opportunities.

-To drive real results, it’s crucial to understand how your AI tool impacts a company’s bottom line. If a business subscribes to your tool, how much time and money will they save? Consider how it can support higher-value work.

-For example, I launched a construction AI company that automates a time-consuming preconstruction process. With our tool, companies save hours of manual work, freeing up their teams to bid on more projects and drive new growth. By delivering such clear value, our business has seen rapid growth and strong adoption.

Kamen Brestnichki, Co-Founder, Martian Lawyers Club

“There is currently a debate going around as to whether generative AI has displaced traditional machine learning. With the underlying models rapidly exploding in size, running a live service that relies on generative AI has unit economics that can quickly break down. While it’s now easier than ever to get to a minimum viable product through third-party generalist models like GPT 4, their prohibitive cost limits their usefulness once at scale.

“It is, therefore, vital for startups that have solidified their use case to invest in smaller, more bespoke models, as those can be trained with an inductive bias towards internal, business-specific data. In that setting, arming your team with old-school (meaning pre-LLM) machine learning knowledge becomes of the utmost importance again.”

Dima Eremin, CEO, BluedotHQ

“I think, just like with any startup, you need to find your niche. It’s important to research the market and find a specific pain point that you can solve with your product or service. This is especially true for AI startups since most people already have access to AI with insanely wide functionality.

“For us, one such pain point was that other note-taking extensions needed a bot to join the call. Users found this awkward and intrusive. And we heard them. We’ve created a note-taker that provides a smoother experience and doesn’t require a bot to “watch” your meetings. Initially, many people told me that maybe we lack a bigger vision or we should somehow make BluedotHQ “sexier.” But the moment we simply focused on what people need was the moment we started to see real growth.”
 

Jim Boswell, CEO, OnPoint Healthcare Partners

“Running an AI startup is all about balancing innovation with real-world impact. AI isn’t just a buzzword; it’s a powerful tool that affects lives, particularly in fields like healthcare, so it’s essential to solve genuine problems rather than using AI for AI’s sake. While it’s tempting to push boundaries, founders should also focus on delivering solutions that are intuitive and trustworthy. Your product can be cutting-edge, but if it isn’t easy to use, you’ll struggle to gain traction. Another critical factor is data quality. AI is only as good as the data it learns from, so it’s worth investing in clean, relevant datasets and processes to keep that quality high.

“Ethics should also be a foundational part of your strategy. With the significant ethical implications AI can carry, especially concerning data, privacy, and fairness, a solid ethics framework is crucial to your brand’s reputation. Flexibility and scalability are also key. AI technology and regulations are constantly evolving, so you’ll want a team and product that can adapt with those changes. Finally, foster a culture of continuous learning. AI moves quickly, and keeping your team engaged with ongoing education and a mindset that embraces change will help you stay competitive.”

Catalin Zorzini, CEO, TutorAI.me

“I feel that the most important thing as founders of AI startups we should think about is our common humanity. AI tools are already so powerful and pervasive, that we might be tempted to maximise every little aspect of managing our business by using some AI feature. I think that moving forward, the most successful entrepreneurs will be the ones wise enough to resist this temptation, and to use their human, personal skills in conjunction with AI tools. The more human touch, the better.”

Swapnil Shnide, CEO, Zeni

“Startups should simplify their products, their customer onboarding, their hiring, and their burn rate, to name just a few key processes. Whenever founders come face-to-face with complex problems, they must ensure their solution adds to simplicity and empowers simpler outputs.

“Simplification also typically leads to leaner teams that get more done with less. By staying focused on simplicity, organisations become more efficient and agile.

“Establishing a simple ideal customer profile (ICP) is one way to navigate customer acquisition challenges with more agility. Companies with a simple ICP have an easier time developing and communicating their value proposition. Consequently, they have a better chance of delivering a message that resonates with the customers who are most likely to convert. When ICPs are complex, they add friction to the sales process and slow growth.

“A simple sales pitch helps prospects to quickly understand the value of the product or service without feeling overwhelmed by jargon or unnecessary details. A simple pitch can also communicate confidence, which helps to position a startup as an authority in its industry that can offer valuable expertise.

“Startups that offer simpler products offer a seamless, frustration-free experience. By reducing issues with UX, regardless of the type of product or service being offered, startups can satisfy customers faster and increase the speed of the decision-making process.”

Erik Schwartz, Chief AI Officer, Tricon Infotech

“AI startups in the UK have had a mix of stories—from standout successes like Synthesia, known for creating AI-generated videos with lifelike avatars for corporate clients, to challenges faced by the likes of Babylon Health, which initially wowed investors but ultimately collapsed, showing just how crucial it is to keep growth expectations realistic.

“Here are some things founders should think about when running an AI startup:

“Solve real problems: It’s easy to be swept up in flashy tech, but make sure your AI solves a genuine market need. It’s the foundation for long-term success.

“Focus on user experience: No matter how smart your AI is, if it’s not easy to use, it won’t take off. Prioritise user-friendly design and seamless integration for your customers.

“Data quality matters: Great AI starts with great data. Founders should secure solid, reliable data sources from day one and stay on top of data management. The quality of your data will directly impact how well your AI performs.

“Hire with vision: You need a team that shares your passion, not just experts with skills. A culture that values creativity and collaboration will fuel your growth.

“Be ethical and transparent: People care more than ever about understanding how AI works. Build ethics and transparency into your tech from the start, so it doesn’t become an afterthought down the road.

“Plan for scalability: Build a strong MVP but make sure your tech can scale as you grow and adapt as AI evolves. Just keep your growth plans realistic—don’t try to run before you can walk.”