BeatpulseLabs, a London-based AI data infrastructure company announced this Monday that it has raised £1.3 million in a pre-seed funding round led by Araya Ventures and Lighthouse Ventures, with additional participation from Alumni Ventures and Avalancha Ventures.
BeatpulseLabs develops high-fidelity training datasets for enterprise artificial intelligence systems. Founded by Jason Rieff and Nikolay Vitanov,the company focuses on transforming existing multimedia content libraries, including speech, music and video, into structured datasets designed for AI model training and fine-tuning.
The company operates through two core services. Its dataset preparation division converts raw multimedia assets into enterprise-grade training data through cleaning, structuring, labelling, validation, enrichment and formatting. Its dataset provision service supplies ready-made and custom rights-cleared datasets for organisations that require high-quality training data without relying on their own archives.
BeatpulseLabs combines licensed datasets, human-in-the-loop annotation and metadata enrichment to produce training data for multimodal AI systems. According to the company, the datasets are designed to improve model accuracy, reduce hallucinations and shorten training times.
BeatpulseLabs reported a tenfold increase in revenue during the first half of 2026, which the company attributed to rising enterprise demand for specialised AI training datasets. The company described the investment as a strategic move to support expansion into additional domains rather than a response to immediate capital requirements.
The growth of multimodal AI systems has increased demand for training data capable of reflecting complex real-world environments.
According to Fortune Business insights, the Multimodal AI Market represents a rapidly evolving segment of the artificial intelligence ecosystem.
The global multimodal AI market size was valued at USD 2.41 billion in 2025. The market is projected to grow from USD 3.32 billion in 2026 to USD 41.95 billion by 2034, exhibiting a CAGR of 37.33% during the forecast period.
BeatpulseLabs positions itself as a data infrastructure provider addressing this need by embedding specialist human expertise and contextual information into AI training datasets.
Investors said the company is addressing a key challenge in enterprise AI by incorporating subject matter expertise and human judgment directly into the datasets used to train advanced AI models.
