An AI matching engine built in 11 weeks

Our client is a Nordic-headquartered influencer marketing platform that plans and runs large-scale social campaigns for major consumer brands across beauty, retail, and on-demand delivery, spanning both Instagram and TikTok. Every campaign starts with a brief and ends in a shortlist of influencers to approach, and that translation was being done entirely by hand. A single brief could take up to a month to become a final shortlist, a hard ceiling on how much new business the company could take on.

They came to gravity9 to build an AI-based recommendation engine, running natively on MongoDB Atlas, that would hold every recommendation to a measurable standard by scoring it against who the company had actually hired in the past, and put the result in front of campaign managers as a live application rather than an offline scoring script.