<\/p>\n
Startup raises $1 million to solve software fragmentation for SMBs.<\/p>\n
Toronto-based Harmix Group<\/a> got its start building an algorithmic search engine for the music industry. Along the way, the team was enamoured by another technical problem: fixing data fragmentation for small businesses using AI.\u00a0<\/p>\n Harmix was an early entrant to the field of multimodal AI, or applying machine learning principles to different media. Founded in 2018, the company patented technology to let users search for images, video, and music using natural-language descriptions.\u00a0<\/p>\n \u201cWhen you have a large library of sometimes millions of compositions, there is no way you would have the time to scroll and to listen to all the previews,\u201d co-founder Nick Shcherban said in an interview. Harmix\u2019s search engine allows users to find the clip they\u2019re looking for just by describing its vibe or uploading a reference file.\u00a0<\/p>\n \u201cWe started developing agents internally to help us, because we had a small team.\u201d<\/p>\n Nazar Ponochevnyi That proved particularly useful for music and television producers. The Harmix API has licensed its tech to media brands such as Red Bull and Sky TV, the latter of which used it to put together the music for Game of Thrones<\/em> spin-off House of the Dragon. <\/em>A recent partnership with music rights management system OpenPlay also put Harmix\u2019s API in the hands of decision-makers at Disney and Warner Brothers.<\/p>\n Co-founder and CEO Nazar Ponochevnyi, who is originally from Ukraine and a graduate researcher at Toronto\u2019s Vector Institute, claimed this part of the company is profitable. But the 20-person company recently found inspiration for a new business direction elsewhere. \u201cOnce we had so much traction with the music, image and video search API, we started developing agents internally to help us, because we had a small team,\u201d Ponochevnyi said.\u00a0<\/p>\n The acceleration of generative AI tools allowed them to kickstart a proactive AI manager (PAM) business, almost like a side-hustle within their own startup. During a regular investor update, the company shared their new agentic interests, and an undisclosed Canadian family office offered them $1 million USD for the new project.\u00a0<\/p>\n RELATED: <\/strong>Backboard.io claims its AI has a better memory than yours<\/strong><\/a><\/p>\n Businesses across sectors are facing pressure to improve efficiency by adopting AI, but smaller enterprises lag behind, according to a federal policy paper<\/a>. But integrating AI without properly connecting different data sources and applications\u2014like Google Drive and Slack\u2014could lead to extra work rather than a streamlined process, Ponochevnyi said. Harmix\u2019s new product aims to help small and medium-sized businesses unify data across these fragmented tools.<\/p>\n Part of Harmix\u2019s decision to build in this new direction came from ecosystem energy.\u00a0<\/p>\n \u201cWith all this hype from OpenClaw, people are really excited,\u201d Ponochevnyi said, referring to the open-source AI agent that performs tasks autonomously<\/a>. \u201cThey want somebody who has expertise integrating those agents into real businesses.\u201d<\/p>\n Harmix will use the funds for hiring engineers and product development as it targets data security certifications and private cloud hosting options. After landing European manufacturing company, Modern Expo, as a client, the company is assembling a cohort of 10 companies it can work with as \u201cdesign partners\u201d to test and improve its new product on their workflows.\u00a0<\/p>\n Feature image courtesy Harmix Group.<\/em><\/p>\n<\/p><\/div>\n\n
Harmix Group<\/cite><\/p><\/blockquote>\n