3 AI Startups in London to Watch

Synthesia – Pioneering AI-Generated Video

Synthesia is a London-based AI startup that has quickly become a global leader in generative AI for video. Founded in 2017 by a team of entrepreneurs and professors, Synthesia’s technology allows users to create realistic video content simply by typing text . How? They leverage AI avatars – digital presenters – who can speak with natural expressions in numerous languages. Instead of filming a live person, you choose one of Synthesia’s avatars (or even create a custom one), input your script, and the AI produces a video of that avatar delivering the lines. It’s essentially video production without cameras or actors.

This innovation is revolutionary for businesses. Companies can generate training videos, marketing messages, or product demos in minutes – no studio time or editing expertise required. Need a how-to video in French and Japanese? Synthesia can produce both with the same avatar, localizing content at the click of a button. The quality of the avatars’ lip-sync and voice is impressively lifelike, thanks to advances in deep learning and years of R&D by Synthesia’s team. The startup’s mission is to make video creation as simple as opening PowerPoint, and they’re well on their way.

Synthesia’s impact is evident in its adoption and backing. It has served 60,000+ customers globally, including big names like Amazon and Johnson & Johnson . Marketers use it for personalized video ads, HR teams use it for onboarding and training videos, and content creators even use it for YouTube or e-learning. The company’s growth attracted major investors – in 2023 they raised a $90M Series C, and by 2025 Synthesia hit a $2.1 billion valuation after a $180M Series D , making it one of London’s AI unicorns. This level of funding and valuation speaks to the market potential they’re tapping: virtually any business or individual that needs video content could be a Synthesia user.

Aside from business use, Synthesia’s tech is being watched closely in the context of media and entertainment. The ability to create digital avatars has implications for film dubbing, gaming, and even resurrecting historical figures virtually. Synthesia maintains ethical guidelines (for example, you can’t create videos of someone without permission) to navigate the deepfake concerns. Their focus, though, is on positive use-cases – enabling content creation that’s faster, more affordable, and more accessible. By being at the forefront of AI video generation, Synthesia firmly puts London on the map in the generative AI arena.

Why watch them: Synthesia is effectively changing how video is made. Just as Canva simplified graphic design, Synthesia is simplifying video production. Their continued innovation (like more customizable avatars, interactivity in videos, etc.) could further disrupt marketing and communications. For businesses, jumping on Synthesia’s capabilities can mean engaging multimedia content with minimal effort. And given their trajectory, an IPO or big partnerships could be on the horizon. They exemplify London’s strength in combining academic research (two of the founders are professors) with commercial execution.

PolyAI – Conversational AI for Customer Service

Another London-founded startup at the forefront of AI is PolyAI, which specializes in building advanced voice assistants for customer service. If you’ve ever called a support line and been frustrated by a clunky phone menu or a bot that doesn’t understand you, PolyAI is the company trying to fix that. Founded in 2017 by a trio of entrepreneurs and researchers from Cambridge University, PolyAI’s platform uses cutting-edge natural language processing to create conversational agents that can actually hold human-like phone conversations .

PolyAI’s voice assistants are deployed by enterprises to answer customer calls, handle inquiries, and carry out tasks just as a human rep would – but available 24/7 and at scale. What sets PolyAI apart is the quality of these interactions. Their AI is designed to understand different accents, phrasing, and even handle clarifications or interruptions gracefully. For example, a bank using PolyAI could have the AI answer with “How can I help you today?” A customer might say, “I lost my card,” and the AI can respond “I’m sorry to hear that. I can help you cancel it and order a replacement. Could you confirm your name and postcode?” – all in a natural, empathetic tone. This goes beyond the rigid scripts of traditional phone bots.

By focusing on natural and engaging voice interactions, PolyAI aims to make the experience seamless . The benefits for businesses are clear: customers get quick service without waiting on hold, and human agents are freed up to handle more complex issues. Companies across hospitality, banking, retail and more have been adopting PolyAI’s solution to modernize their call centers. One notable deployment is with hospitality chains where their AI concierge handles guest requests like booking changes or room service orders over the phone, improving response times and consistency.

PolyAI’s progress is reflected in its funding. They secured a €10.7M Series A in 2019 and more recently a $50M Series C in 2024 , indicating strong investor belief in their tech and market. The demand for automated customer service skyrocketed during the pandemic and continues as businesses seek efficiency, so PolyAI finds itself in a sweet spot. They’re now one of the leading conversational AI startups globally, not just in London.

Why watch them: PolyAI is demonstrating that AI can handle voice conversations that were once thought too complex for machines. As voice AI becomes more ubiquitous (think Alexa, Siri, etc., but for business-specific tasks), PolyAI is positioned as a specialist delivering bespoke solutions to enterprises. If they continue improving, calling a company and speaking to an AI that you don’t even realize is an AI could become normal – and PolyAI would be a big reason why. They are a standout in London’s growing NLP (natural language processing) scene and show how the city isn’t just about fintech or healthtech, but also leading in AI for customer experience.

Tractable – AI for Insurance Claims and Disaster Recovery

Our third London AI startup to watch is Tractable, founded in 2014. Tractable applies AI to a very practical (and financially significant) problem: assessing damage and processing insurance claims. Using computer vision, Tractable’s algorithms can analyze photos of damage (like a car after an accident or a house after a storm) and instantly estimate repair costs, detect severity, and more . Essentially, they train AI models on millions of images of wrecked cars, for example, so that given new photos, the AI can say “This car’s front bumper is damaged, estimated repair $800, likely needs replacement of these parts.”

Traditionally, insurance companies send human adjusters to inspect damage – a time-consuming and sometimes subjective process. Tractable’s solution can do it in minutes from just digital photos uploaded via an app. This speeds up claim approvals dramatically – some insurers using Tractable have been able to settle claims within the same day of the incident. Customers benefit by getting paid faster, and insurers benefit by handling more claims more efficiently (and catching fraud or inconsistencies better through AI’s keen eye).

Tractable’s early focus was on auto insurance, partnering with major global insurers like GEICO and others . If you’ve had a fender-bender and used an app to submit photos to your insurer, there’s a chance Tractable’s AI was behind the scenes making sense of those images. The startup has since expanded its AI to other areas like property damage. After natural disasters, for example, their tech can analyze aerial imagery of an affected region to triage which homes are most damaged and need urgent attention. This is incredibly valuable when thousands of buildings might be hit by a hurricane – AI can survey the damage far faster than humans on the ground.

Tractable’s success is evident in its funding and growth. They raised a $65M Series E round led by SoftBank’s Vision Fund in 2021, which put them into “unicorn” status with a valuation over $1B . By 2025 they’ve continued to grow globally, with offices in North America, Asia, and Europe serving insurers and manufacturers (they also use their AI in car recycling to identify reusable parts). The company stands out because it brought AI to a very traditional industry (insurance) and proved its worth. It also shows off one of London’s strengths – the intersection of fintech/insurtech expertise with AI talent.

Why watch them: Tractable is a great example of an AI startup solving real, tangible problems and not just building tech for tech’s sake. As climate change potentially increases the frequency of extreme weather events, tools like Tractable’s could become vital for disaster response and recovery, beyond just insurance claims. They are continuously improving their AI accuracy and expanding to new types of damage assessment. Moreover, the partnership with big insurers means Tractable’s tech could become a standard in the industry – which is a huge win for a startup. Keep an eye on their tech being used in new ways, such as helping assess and price used cars, or evaluating infrastructure damage in public sector projects. Tractable’s journey from a small London founding team to a global insurtech leader is inspiring and indicative of the practical innovation coming out of the UK.

‍