\n<\/p>\n
The majority of companies struggle to extract value from their data. Several years ago,\u00a0Forrester\u00a0reported that between 60% and 73% of data belonging to the average business goes unused for analytics<\/a>. That\u2019s because the data\u2019s siloed or otherwise pigeonholed by technical and security considerations, making it difficult \u2014 if not impossible \u2014 to apply analytical tools.<\/p>\n Anna Pojawis and Tyler Maran, engineers who previously did stints at Y Combinator-backed startups Hightouch (a data-syncing platform) and Fair Square (a health insurance tool), were inspired to try their hands at solving the data value problem after discovering that many companies had been \u201clocked out\u201d of analytics strategies due to the engineering roadblocks. <\/p>\n \u201cWe\u2019ve found that a significant part of the market, especially those in regulated industries like healthcare and finance,\u201d have struggled with data analytics, Maran told TechCrunch. \u201cThe majority of corporate data doesn\u2019t fit into a database today; it\u2019s sales calls, documents, Slack messages and so on. And, given the scale of these companies, off-the-shelf data models are typically not sufficient.\u201d<\/p>\n So Pojawis and Maran founded OmniAI<\/a>, a set of tools that transform unstructured enterprise data into something that data analytics apps and AI can understand.<\/p>\n OmniAI syncs with a company\u2019s data storage services and databases (e.g., Snowflake, MongoDB, etc.), preps the data within and allows companies to run the model of their choice \u2014 for example, a large language model \u2014 on the data. OmniAI performs all of its work in the company\u2019s cloud, OmniAI\u2019s private cloud or on-premises environments, delivering ostensibly improved security, according to Maran.<\/p>\n \u201cWe believe that large language models will become essential to a company\u2019s infrastructure in the next decade, and having everything hosted in one place just makes sense,\u201d Maran said. <\/p>\n Out of the box, OmniAI offers integrations with models, including Meta\u2019s Llama 3<\/a>, Anthropic\u2019s Claude<\/a>, Mistral\u2019s Mistral Large<\/a> and Amazon\u2019s AWS Titan for use cases like automatically redacting sensitive information from data and generally building AI-powered applications. Customers sign a software-as-a-service contract with OmniAI to enable management of models on their infrastructure.<\/p>\n It\u2019s early days. But Omni, which recently closed a $3.2 million seed round led by FundersClub at a $30 million valuation, claims to have 10 customers already, including Klaviyo and Carrefour. Annual recurring revenue is on track to reach $1 million by 2025, Maran said.<\/p>\n \u201cWe\u2019re an incredibly lean team in a fast-growing industry,\u201d Maran said. \u201cOur bet is that, over time, companies will opt for running models alongside their existing infrastructure, and model providers will focus more on licensing model weights to existing cloud providers.\u201d<\/p>\n<\/div>\n