Advertise on podcast: Riding Unicorns: Venture Capital | Entrepreneurship | Technology

AI is creating extraordinary new demand for electricity. But what happens when the grid can't keep up? In this episode of Riding Unicorns, James and Hector sit down with Karl Bach, Co-Founder & CEO of Axle Energy, the energy technology company building a virtual power plant from hundreds of thousands of electric vehicles, batteries and heat pumps. Rather than relying solely on large power stations to balance electricity supply and demand, Axle connects distributed devices across homes and businesses and intelligently shifts when they consume energy. Today, the company manages more than two gigawatts of capacity, roughly equivalent to the capacity of a nuclear reactor. The timing is becoming increasingly important. Renewable energy creates more variability in electricity supply, while AI and the rapid expansion of data centres are creating new sources of demand. In the UK, Karl says a new data centre can face a grid connection wait of around 10 years. Axle's technology can help provide flexibility to the electricity system in weeks rather than waiting years for new physical infrastructure to be built. Karl also shares the realities of building in a heavily regulated industry. Axle spent years helping push through regulatory changes required for distributed devices to participate properly in energy markets, while gradually building from its first EV charger partnerships into a platform managing hundreds of thousands of devices. Topics Covered • How virtual power plants work • Why EVs, batteries and heat pumps can help balance the electricity grid • Building more than two gigawatts of capacity from distributed devices • Why renewable energy increases the need for grid flexibility • The impact of AI and data centres on electricity demand • Why connecting a new UK data centre to the grid can take around a decade • Building a startup inside a heavily regulated industry • How Axle won its first customers before it had a product • Creating recurring revenue for EV charger manufacturers • Karl's experience leading growth at Bulb • What Bulb's collapse taught him about building in the physical world • Balancing startup speed with the consequences of getting things wrong • Hiring exceptional people inside a highly technical industry • Why Karl doesn't believe energy will ever become free • Building a company around a problem you genuinely care about Karl also discusses his experience at Bulb, once one of Europe's fastest-growing companies, and how its eventual collapse following the energy crisis shaped his understanding of the risks that come with building businesses exposed to the physical world. This is a conversation about one of the fundamental constraints on the next wave of technology: energy. How we produce it, how we move it, and how software can make the infrastructure we already have significantly more efficient.

What happens when companies deploy AI agents before fixing the data those agents rely on? In this episode of Riding Unicorns, James and Hector sit down with Toby Mather, Co-Founder of Rig, the AI-native data infrastructure company building the data layer that allows humans and AI agents to work from the same trusted context. Rig started as a "data brain" sitting on top of platforms like Snowflake and BigQuery, making sense of fragmented company data so teams could use tools like Claude and Cursor to build applications, automations and analytics. It has since expanded into an end-to-end data stack covering ingestion, warehousing and context. Toby explains why LLMs alone aren't the answer to enterprise data. While language models are excellent at understanding context, businesses still need deterministic systems like SQL and structured tables when they want accurate answers from their data. Rig's thesis is that AI should make sense of the complexity, but the underlying calculations still need to be grounded in reliable data. The conversation also explores the rush towards autonomous AI agents. Toby argues that many companies are trying to run before they can walk: connecting agents to dozens of fragmented systems without first creating the clean, consistent data infrastructure they need. Get the context right, and the journey from human-in-the-loop to autonomous agents becomes much easier. Topics Covered • Why AI agents need better data infrastructure • Building a "data brain" for humans and AI agents • Why LLMs aren't always good at data reasoning • Why 55-year-old SQL and tables still matter in the AI era • Moving from human-in-the-loop to autonomous agents • How Rig uses Rig to run its own business • Building agentic workflows across CRM and customer success • Toby's lessons from building and selling Lingumi • Why founders should choose enormous markets • Landing major enterprise customers as an early-stage startup • The PULL framework for B2B sales • How to sell enterprise software without becoming a services business • Competing with incumbents like Snowflake and Databricks • Balancing customer feedback with founder-led product intuition • Why agentic customer success could be a major opportunity Toby also shares some of the more surprising ways customers are already using Rig, from identifying overpayments hidden across thousands of invoices to building AI-powered customer success systems that identify problems and proactively contact customers. This is a conversation about the infrastructure beneath the AI agent revolution: getting company data into the right shape, giving AI the context it needs and building the foundations required for truly autonomous work.

What happens when AI moves beyond the digital world and starts transforming the companies that manufacture the things around us? In this episode of Riding Unicorns, James and Hector sit down with Robin Van Aeken, Co-Founder & CEO of Magentic, the AI company building digital workers for some of the world's largest manufacturers. Magentic's AI agents, or "mages", work alongside procurement and supply chain teams at companies including Siemens and Coca-Cola Europacific Partners. They connect with existing systems, analyse huge volumes of procurement data and identify opportunities to make supply chains cheaper, faster and more resilient. The economics can be significant. A manufacturer spending $10 billion a year on procurement might save $50 million by making its team 50% more efficient. But improving procurement costs by just 1-2% could generate $100-200 million in savings. Robin explains why this makes procurement one of the most compelling applications for enterprise AI. The conversation also explores how Magentic deploys AI agents inside complex enterprises. Rather than expecting autonomy from day one, its digital workers effectively begin as "interns", learning the organisation, its systems and its data before gradually earning the trust to take actions and eventually run significant parts of procurement and supply chain workflows. Topics Covered • Why the next major AI opportunity could be in the physical economy • Building AI digital workers for procurement and supply chains • Why AI can create more value by reducing procurement spend than headcount • How human-in-the-loop feedback helps agents develop domain expertise • Using AI across complex ERP systems like SAP and Oracle • Building trust with some of the world's largest manufacturers • Data privacy, sovereignty and deploying AI inside sensitive organisations • How Magentic's "mages" progress from interns to autonomous digital workers • Why AI agents increasingly interact with software like humans • Robin's journey from McKinsey to founding Magentic • What his co-founder's experience at OpenAI taught them about the physical economy • Building an enterprise AI company backed by Sequoia Capital • How to construct the right investor syndicate beyond simply raising capital • Why faster, cheaper supply chains could ultimately reduce the cost of physical goods Robin also explains why the explosion in AI infrastructure is creating a huge new challenge for the physical economy. Data centres, chips, power infrastructure, cooling and other components all depend on manufacturing and supply chains that weren't designed for today's speed of demand. This is a conversation about bringing AI into the physical economy: deploying agents inside complex enterprises, making global supply chains more efficient and using software to unlock potentially hundreds of millions of dollars in savings.

What happens to software engineering when almost every line of code is written by AI? In this episode of Riding Unicorns, James sits down with Mirko Novakovic, Founder & CEO of Dash0, the AI-native observability company helping engineering teams understand, monitor and fix problems in increasingly complex software systems. Dash0 has grown at remarkable speed, 4x-ing ARR in six months and recently raising a $110m Series B. For Mirko, now on his third company, that growth hasn't come from a single breakthrough. His philosophy is much simpler: build the fundamentals brick by brick, continuously improve them, and let the effects compound.  The conversation explores how AI is fundamentally changing software development. At Dash0, almost no code is now written manually, with engineers increasingly using natural language to create software and focusing their time on architecture, review and quality. Mirko believes the next step is automating the entire journey from prompt to production.  We also discuss what it takes to scale a venture-backed company aggressively. Mirko explains why founders need to recognise when they have momentum and "step on the gas", why revenue growth is the metric that matters most in the early stages, and why sometimes you need to shorten your runway to capture the opportunity in front of you.  Topics Covered • Why AI-generated code creates a bigger need for observability  • Growing Dash0 from launch to unicorn at exceptional speed  • Why there is no "secret sauce" to building a great software company  • How small improvements compound into significant growth  • Knowing when to step on the gas and scale aggressively  • Why revenue growth matters more than efficiency at the early stage  • Saying no to $10m+ customer opportunities when you're not ready  • Why the first engineers you hire must be exceptional  • How AI is changing the role of software engineers  • Why almost every line of code at Dash0 is now AI-generated  • Building an AI-first organisation without restrictive procurement processes  • Why Dash0 insists presentations are generated entirely by AI  • Developers spending up to $20,000 per month on AI tools  • Turning internal AI tools into new products  • Why AI-native companies are documenting and systemising more of their knowledge Mirko also shares his approach to hiring, why great people attract other great people, and how his experience building previous companies has shaped the speed and ambition behind Dash0. This is a conversation about building at machine speed: using AI across engineering and operations, knowing when to take risk, and creating the organisational momentum required to build a category-defining software company.

How do you build a global AI company without burning hundreds of millions of dollars? In this episode of Riding Unicorns, James sits down with Krish Ramineni, Co-Founder & CEO of Fireflies.ai, the AI meeting assistant used by millions of people around the world. Fireflies started building with AI long before the current boom. After struggling to raise capital, Krish and his co-founder became obsessed with reaching "ramen profitability", keeping costs low and controlling their own destiny. They eventually raised venture funding, but haven't touched the capital from their Series A, reaching profitability in 2023 while continuing to grow rapidly. The conversation explores how Fireflies created the AI notetaker category and engineered distribution directly into the product. Rather than relying on traditional enterprise sales, Fireflies focused on self-service adoption, word of mouth and a viral loop where every meeting introduced the product to new users. This helped Fireflies spread to more than 100 countries within its first year. Krish also explains where Fireflies goes next: beyond meeting notes towards AI that understands conversations across meetings, email and Slack, identifies what needs to happen and eventually completes work while conversations are still taking place. Topics Covered • Building an AI company before the AI boom  • Why Fireflies prioritised profitability from the beginning  • Creating the AI notetaker category  • Engineering virality and word of mouth into the product  • Spending almost nothing on marketing until reaching tens of millions in revenue  • Going horizontal when VCs advised Fireflies to focus on enterprise sales  • Scaling a self-service product across 100+ countries  • Building infrastructure for millions of users  • Privacy, security and customer data in AI  • Expanding from meeting notes into email, Slack and AI agents  • Building a profitable company while maintaining venture-style growth  • Krish's first-principles approach to leadership  • Why iteration speed compounds into competitive advantage  • The opportunity in open-source AI and inference infrastructure Krish also shares why he runs Fireflies with an unusually flat organisation and more than 30 direct reports, and why he believes startups win by iterating faster than their competitors. This is a conversation about building an AI company differently: staying lean, engineering distribution into the product, obsessing over economics and surviving long enough for the market to catch up with your vision.

How do you build a consumer product that people genuinely love, while still making the economics work? In this episode of Riding Unicorns, James sits down with Gyve Safavi, Co-Founder & CEO of SURI, the award-winning electric toothbrush company combining design, performance and sustainability. Before founding SURI, Gyve spent eight years at Procter & Gamble, ultimately leading brand management for the £500m Gillette Female Shave Care business across EMEA. He later worked at AKQA and WPP, including alongside Sir Martin Sorrell, before moving into startups. The conversation explores how those experiences translated into building SURI and what it really takes to create a modern consumer hardware brand. Gyve explains how SURI was built around three principles: exceptional design, high-performance cleaning and sustainability. He breaks down how the team used consumer research to determine pricing before launch, why gross margin needs to be engineered into the business early, and how relentless negotiation with suppliers has improved unit economics as the company has scaled. We also explore customer acquisition, manufacturing, co-founder dynamics and how SURI is using AI internally to build a more efficient organisation. Topics Covered • Gyve’s journey from P&G, AKQA and WPP into entrepreneurship • Building SURI around design, performance and sustainability • How to price a consumer product before launch • The four questions SURI used to identify its optimal price point • Why gross margin can make or break a consumer startup • Negotiating with manufacturers as volumes grow • Reducing customer acquisition costs through better execution • Why relationships can create an unfair advantage for startups • Winning a month of free advertising at London’s Outernet simply by asking • Co-founder dynamics and SURI’s early “bitching sessions” • Why founders should listen to customers more than investors • Learning to handle setbacks and celebrate wins • How SURI is using AI to operate more efficiently • Why AI could make this one of the best periods to build a consumer company Gyve also shares his future unicorn predictions and explains why his ideal dinner would include Jony Ive, Jeff Bezos and Mick Jagger. This is a practical conversation about building a modern consumer brand: creating a product people want, getting the economics right and building an organisation capable of scaling with it.

What happens when you hear your own hit song playing in public, but never receive the royalty? In this episode of Riding Unicorns, James sits down with Ryan Edwards, Founder & CEO of Audoo, the music technology company rebuilding how public performance royalties are measured and distributed. The idea for Audoo came from Ryan’s own experience as a musician. After hearing one of his songs playing in a London department store and discovering there was effectively no reliable way to track the royalty, he started investigating how the system actually worked. What he found was an industry still relying heavily on proxy data, manual surveys and people physically visiting venues with clipboards. Audoo’s answer is the Audio Meter, a small connected device that identifies music being played in public venues without recording or storing conversations. The resulting data allows collecting societies to understand what was actually played and distribute royalties more accurately to artists and rights holders. Today, Audoo operates across 16 countries, with its technology deployed in environments ranging from cafés and retailers to gyms and leisure centres. Ryan shares how he built the first prototype on a Raspberry Pi at his kitchen table, validated the idea with the former Chairman of PRS for Music, and launched the first real-world deployment on the other side of the world in Australia during COVID. We also discuss the unusual group of investors Audoo has attracted along the way, including Björn Ulvaeus of ABBA, Paul McCartney, Elton John, Adam Clayton of U2 and Adele, and how those relationships have helped build credibility and open doors across the music industry. Topics Covered • Why the public performance royalty system needed rebuilding • The personal experience that led Ryan to start Audoo • Building the first Audio Meter prototype on a Raspberry Pi • Why privacy had to be designed into the technology from day one • The challenges of combining software, data and physical hardware • Landing Audoo’s first major deployment in Australia during COVID • Why venues are willing to install Audio Meters • How better music recognition data changes royalty distributions • Raising capital from some of the biggest names in music • Expanding Audoo across 16 countries and into the Middle East • What Ryan has learned from eight years as a founder • Why founders need to celebrate wins and know when something can wait until Monday This is a conversation about spotting an antiquated system, using technology to rebuild it from first principles, and creating the data infrastructure that could finally ensure artists are paid accurately when their music is played.

What does it take to build a frontier AI model lab in a world dominated by OpenAI, Anthropic and Google? In this episode of Riding Unicorns, James and Hector sit down with Yang Li, Co-Founder & CEO of Cosine, one of the UK's leading frontier AI companies building sovereign AI models for enterprise and government. Cosine began by fine-tuning foundation models for software engineering before evolving into a full frontier AI lab, training its own models from scratch. Today, the company works with organisations across defence, financial services and critical national infrastructure, deploying AI securely inside highly regulated environments. The conversation explores why sovereign AI has become a strategic priority, how smaller model labs can compete with the biggest players, and what the future of enterprise AI will look like. Yang also shares how AI is changing software engineering, why synthetic data has become a competitive advantage, and the lessons he's learned building one of Europe's most ambitious AI companies. Topics Covered • How Cosine evolved from AI coding tools into a frontier AI model lab  • Why sovereign AI is becoming critical for governments and enterprises  • Competing with OpenAI and Anthropic without billions in funding  • Building AI models using synthetic data and post-training techniques  • Why defence, banking and healthcare are driving enterprise AI adoption  • The future of software engineering in the age of AI coding agents  • Air-gapped AI, on-premise deployment and enterprise security  • The geopolitical race for AI infrastructure and compute sovereignty  • Hiring world-class AI talent and building high-performance teams  • Founder lessons on conviction, resilience and staying alive long enough to win This is a conversation about the future of frontier AI, the growing importance of sovereign technology, and what it takes to build an AI company that competes on the global stage.

What happens when computing reaches the limits of classical physics? In this episode of Riding Unicorns, James and Hector sit down with Ashley Montanaro, Co-Founder & CEO of Phasecraft, one of Europe's leading quantum computing companies. Backed by Episode 1, Phasecraft is building the software and algorithms that will unlock the commercial potential of quantum computers. While much of the industry focuses on hardware, Ashley explains why the real breakthroughs will come from the software layer, and why better algorithms could be worth orders of magnitude more than incremental hardware improvements. The conversation explores what quantum computing actually is, why "quantum advantage" could become one of the defining technology milestones of the next decade, and how quantum and AI will work together to accelerate scientific discovery. Ashley also shares what it's really like building a deep tech company, how non-dilutive funding helped Phasecraft grow, and why the next generation of computing could transform industries ranging from energy and pharmaceuticals to materials science and cybersecurity. Topics Covered • What quantum computing actually is and why it matters  • Why software may create more value than quantum hardware  • The race towards practical quantum advantage  • How quantum algorithms unlock problems classical computers cannot solve  • The future of materials discovery, batteries, pharmaceuticals and chemistry  • Why AI and quantum computing will reinforce each other  • The cybersecurity implications of post-quantum computing  • Building a deep tech company with long development cycles  • Raising over £5m in non-dilutive funding through grants and government programmes  • Commercialising frontier research before the market fully exists  • Lessons from building one of Europe's leading quantum software companies  • Why radically new computing architectures could define the next generation of technology This is a conversation about one of the most important technological revolutions still to come, why quantum computing is moving from theory to commercial reality, and what it takes to build a company at the frontier of science and software.

How do you build AI for one of the world's largest, most valuable, and least digitised industries? In this episode of Riding Unicorns, James and Hector sit down with Dr. Kate Jarvis, CEO & Co-Founder of Fifth Dimension, the AI company transforming how institutional real estate investors make decisions. With a PhD from Stanford in machine learning and linguistics, Kate has spent her career applying AI to complex industries. Today, Fifth Dimension is building what she describes as the Bloomberg Terminal for real assets, helping some of the world's largest property investors replace spreadsheets, PDFs and manual workflows with AI-powered decision intelligence. The conversation explores why commercial real estate has remained decades behind other financial markets, how AI is finally unlocking the sector, and what it takes to build enterprise software for one of the world's most relationship-driven industries. Kate also shares candid insights on fundraising in today's AI market, founder-led sales, pricing AI products, and why Europe may have a unique advantage in the next wave of vertical AI companies. Topics Covered • Why commercial real estate remains one of the world's least digitised industries  • Building the Bloomberg Terminal for institutional property investors  • How AI transforms underwriting, portfolio management and investment decisions  • Why decision intelligence matters more than simply collecting more data  • The challenge of changing behaviour in relationship-driven industries  • Founder-led sales and scaling enterprise go-to-market teams  • Pricing AI products while managing token costs and unit economics  • Building vertical AI companies versus foundation models  • The differences between fundraising in Europe and the US  • Why Europe could become the leader in vertical AI applications  • The realities of raising venture capital as an AI founder  • Founder resilience, conviction and building through uncertainty This is a conversation about applying AI where it creates the greatest economic value, why enterprise transformation is ultimately about people rather than technology, and what it takes to modernise one of the world's biggest industries.

Laura Modiano leads Founder Experience at OpenAI, a newly created function designed to help startups build, grow, and scale on the world's leading AI platform. Having previously worked with startups at AWS and now at OpenAI, Laura sits at the intersection of founders, technology, and the future of company building. In this episode, Laura shares how AI is fundamentally changing the startup journey, why founders can now build companies faster than ever before, and what OpenAI is doing to support the next generation of builders from day one. We discuss: • Why OpenAI created the Founder Experience team • How startup building is changing in the AI era • What founders need from AI platforms beyond infrastructure • The difference between building a feature and building a company • Why AI-native founders are moving faster than ever before • How OpenAI gathers feedback directly from founders • The rise of non-technical founders building technical products • Why anyone can now learn, build and create with AI tools • How Laura uses Codex to automate workflows and increase productivity • The future relationship between startups and AI platforms • Lessons from AWS, healthcare, startups and OpenAI • What makes great founders stand out in a rapidly changing market Laura also shares her thoughts on the future of AI applications, the opportunities for founders building on top of foundation models, and why curiosity, speed and continuous learning are becoming the most valuable skills in technology. This is a conversation about the future of entrepreneurship, the democratisation of software creation, and what happens when anyone with an idea can build a company.

What does it take to build an AI company a decade before the world realises AI is the future? In this episode of Riding Unicorns, James and Hector sit down with Barney Hussey-Yeo, Founder & CEO of Cleo, the AI financial assistant helping millions of consumers make smarter money decisions. Barney started building Cleo years before ChatGPT brought AI into the mainstream. Today, Cleo has grown into one of the UK's leading fintech success stories, reaching $400M ARR and becoming a category-defining consumer AI company. The conversation explores the realities of scaling from zero revenue to hundreds of millions in ARR, surviving the shift from the ZIRP era to COVID, and building a product-led company that endured long before AI became fashionable. Barney also shares how AI is transforming the way Cleo operates internally, why every executive team member is now shipping code, and how founder behaviour, company culture, and execution speed are changing in the age of AI. Topics Covered: • Building an AI company before the AI boom  • The origins of Cleo and the vision for an AI financial assistant  • Growing from $0 to $400M ARR  • Why the first $100M is the hardest milestone  • Surviving COVID and finding product-market fit  • Building data moats and defensibility in AI  • Why AI has transformed customer acquisition  • How Cleo uses AI internally across engineering and operations  • The rise of AI-native companies and the future of work  • Why some employees thrive with AI and others fall behind  • The best AI tools Barney is using today  • Advice for founders on fundraising, execution, and building great products  • The next generation of AI startups and future unicorn predictions A candid conversation about ambition, execution, AI, and what it really takes to build a generational technology company.

What happens when software adapts to humans, rather than humans adapting to software? In this episode of Riding Unicorns, James and Hector sit down with Jonathan Sanders, Founder & CEO of Light, the AI-native finance platform redefining how businesses manage money. Jonathan's journey spans investment banking, consulting, Pleo, founding Juni, and now building Light. Along the way, he has developed a unique perspective on technology, AI, and what the next generation of business software will look like. The conversation explores Jonathan's concept of "organic software" where AI agents learn, adapt, and operate alongside teams, reducing complexity and transforming finance from a reporting function into a strategic intelligence layer. Jonathan shares real-world examples of AI agents managing company spend, monitoring compliance, automating finance workflows, and generating insights that would have been impossible just a few years ago. The discussion also covers founder psychology, building through uncertainty, AI-first organisations, and how leadership teams must evolve in a world where software development, product design, and operations are increasingly powered by AI. Topics Covered: • Jonathan's journey from finance and consulting into startups • Why founding a company is one of the hardest ways to create change • The concept of "organic software" and AI-native business tools • How AI agents learn from company behaviour and improve over time • Building finance software that tells stories instead of generating reports • Real-world examples of autonomous finance workflows • Why AI is changing the role of CFOs and finance teams • How Light uses AI internally across product, engineering, and operations • The rise of AI-first organisations and new ways of working • Founder resilience, stress, and the realities of building companies • Why software development is changing faster than most people realise • The future of business operations in an AI-native world This is a conversation about the next generation of software, the future of work, and what happens when AI becomes a true operating system for businesses.

Steve Domin, Founder & CEO of Duffel, the company rebuilding the infrastructure layer of the global travel industry. Duffel is taking on one of the most complex and outdated sectors, creating modern, developer-first APIs for flights, hotels, and more. Backed by investors including Index Ventures and Benchmark, the company has raised over $50M to transform how travel is bought and sold. Steve shares the journey of building Duffel from scratch, including:  Why the travel industry is fundamentally broken  The insight that led to Duffel’s API-first approach  Building deep supply-side integrations with airlines and incumbents  Navigating COVID when travel demand dropped to zero  The painful reality of scaling, resetting, and rebuilding product-market fit  Where long-term defensibility comes from in complex infrastructure businesses We also explore:  What makes companies like GoCardless “talent factories”  How AI is changing how modern engineering teams are built  The future of smaller, highly efficient companies  Steve’s view on the next generation of AI-native products This is a masterclass in persistence, infrastructure thinking, and building through uncertainty.

Mahdi Yahya, Founder and CEO of Ori, is now part of Radiant following its partnership with Brookfield. Mahdi is building at the centre of one of the most important shifts in technology: the global race to scale AI infrastructure. We cover:  From telecoms to data centres: lessons from building infrastructure businesses from age 20  Why Ori chose the harder path of building for a future that did not yet exist  What “AI infrastructure” actually means, from GPU clusters to orchestration software  The Radiant model: combining capital, power, compute and software into a single platform  Why AI is becoming sovereign infrastructure and what that means for governments  Where value will accrue in the AI stack across hardware, infrastructure and applications  The importance of vertical integration and marginal gains in a capital-intensive market  Two pivotal decisions that shaped the company, including early acquisition and moving into physical infrastructure  Founder resilience and the simple rule: stay alive Mahdi also shares his perspective on the next wave of AI companies, the opportunity in inference, and why robotics could be the next breakout category. This is a conversation about building at civilisational scale and what it takes to operate where technology, capital and infrastructure collide.