Lightning seeks to simplify artificial intelligence management.
Lightning AI, a development platform based on the open framework PyTorch Lightning, has successfully raised new funds in a recent funding round.
The development and implementation of artificial intelligence (AI) continue to be a considerable challenge for many organizations, despite AI being a hot topic today. A recent study by Boston Consulting Group indicates that 74% of companies are struggling to extract value from their AI investments. William Falcon, creator of PyTorch Lightning, a widely used open-source AI framework, points out that one of the most common mistakes companies make is underestimating the amount of effort required to orchestrate AI.
Falcon compares building an AI platform to creating a communication tool like Slack, arguing that it is a complex, costly process that is not essential for the business. According to him, the real value for companies lies in their data, domain knowledge, and unique models, rather than in maintaining the AI infrastructure.
With experience as a former Navy Seals trainer and intern at Facebook AI Research, Falcon began developing PyTorch Lightning during his time as a student at Columbia. This framework provides a high-level interface for the PyTorch AI library, simplifying the management and configuration of AI systems. After leaving his Ph.D. program at NYU, Falcon joined Luis Capelo, former head of data products at Forbes, with the goal of commercializing PyTorch Lightning. Together, they founded Lightning AI, which applies services and tools directed at businesses based on the open-source framework.
Lightning AI enables developers to train and deploy models at scale, a task that previously required large teams. The company handles typically complicated tasks, such as distributing AI workloads and providing the infrastructure needed to evaluate and train models. Its flagship product, AI Studios, allows clients to fine-tune and run AI models in their preferred cloud environments. Additionally, businesses have the option of using Lightning AI to host AI-driven applications on private cloud infrastructures or in their own data centers. The pricing model is pay-as-you-go, with a free option that includes 22 hours of GPU usage per month.
Falcon explains that the goal of Lightning AI is to make AI development as intuitive as using an iPhone. The platform has allowed researchers at Columbia to conduct hundreds of experiments within a 12-hour span. He also notes that many of the world's leading AI products have been developed using Lightning, such as Nvidia's NeMo models and Stability AI's Stable Diffusion.
Lightning AI is experiencing significant growth, with over 230,000 AI developers and 3,200 organizations using its platform. The company recently raised $50 million in a funding round, bringing its total capital to $103 million. Aiming to attract new clients, including from the government sector, and expand the Lightning platform into new markets, the New York-based company, which has a team of 50 people, plans to use this funding to strengthen its market presence.
Despite competition from other companies offering similar services, such as Comet, Galileo, and Weights & Biases, Falcon believes there is enough room in the market for all players. According to forecasts by Fortune Business Insights, the machine learning operations sector, in which Lightning AI is positioned, could reach a value of approximately $13 billion by 2030. Falcon also expects Lightning AI to achieve between $10 million and $20 million in annual recurring revenue by the end of next year and to reach profitability shortly thereafter.
