We built Fanlytiq because we kept watching good video teams get hurt by bad timing.
— Priya Nambiar, CEO & Co-Founder
From Cirrus Media to a pre-publish tool
Before founding Fanlytiq in 2023, Priya Nambiar spent five years in digital media analytics at Cirrus Media, a streaming platform operating mid-market content series across AVOD and SVOD tiers.
The pattern that drove her out was consistent: teams would publish a video on Monday, watch the algorithm's recommendation window open and close by Wednesday, and only receive meaningful engagement data Thursday. The talent was good. The content was good. The timing was wrong.
The decision to build in Los Angeles was deliberate. The studio and streaming ecosystem concentrated in LA represents the densest cluster of Fanlytiq's core customer — content teams and independent streaming operations that publish on volume and live and die by the first 72 hours.
"The teams weren't failing because they made bad content. They were failing because the only signal they had was a 48-hour-delayed post-mortem. We wanted to move that window — all the way to before the video goes live."
Three people who built the thing they needed
Five years in digital media analytics at Cirrus Media. Built Fanlytiq to move the engagement intelligence window from 72 hours post-publish to before the algorithm sees the video.
ML engineer background. Previously built viewer recommendation systems at Nexlode Analytics. Designed the format-specific prediction architecture that lets Fanlytiq score YouTube long-form, Shorts, Reels, and episodic series with separate models — not a single averaged one.
Content strategy and product background, previously at Wavemark Studios. Shapes Fanlytiq's product philosophy: every output is an action, not a chart.
Three operating principles
We never give you "your video scored 72%." We give you the exact timestamp and the likely cause. A score without a location is not actionable.
The only useful moment to know about a problem is before the audience sees it. Post-publish data is a historical record. Pre-publish data is a decision tool.
Every output from Fanlytiq is an action: cut this segment, swap this thumbnail, restructure this sequence. If the output doesn't suggest a specific edit, it isn't useful.