Luiz Alberto Tessarolli

Hey StartupBase, I'm Luiz, solo founder and engineer behind Recrutador. I built it after years of watching the best-talking candidate win the interview while the best person for the role quietly lost. Selection science solved this 80+ years ago (structured interviews far outpredict gut-feel ones), but the method stayed locked in academia and enterprise tooling. Recrutador runs the whole hiring process on one ruler: define the criteria, triage resumes against them, get real-time guidance during the live interview, and end with a report that anchors every rating in the candidate's own words. It assists the interviewer, it never decides, and audio is never stored. Would love your feedback, especially from anyone who hires and has felt that gap.

The saturation-level signal is the part that actually matters. Anyone can surface a "trending" product, but by the time it's trending in Facebook ad intelligence the margin is usually already gone. If you can show me how saturated a product is (how many stores already running it, ad-spend trend over time), that's the difference between a real winner and a graveyard everyone's piling into. The competitor-strategy view is a strong complement. One question: how fresh is the ad data? Real-time genuinely matters here, since dropshipping windows close in days, not weeks.

Pulling the site straight from existing listings and reviews is the right wedge: the businesses that need this most (local services, restaurants) are exactly the ones who will never sit down and write copy. The review angle is underrated too, a page built from real customer quotes converts better than founder-written marketing almost every time. My one worry is freshness, listings and hours and reviews all drift. Does it auto-resync when a new 5-star lands or hours change, or is it a one-time generate? That recurring sync is what would make me keep paying past month one.

The "highlight where viewers get confused" angle is the smart inversion, most analytics tell you where people dropped off but never why, and the why is the whole game. The hard part is signal quality: a pause or a rewatch can mean confusion or genuine interest, so how you disambiguate intent will make or break the insight. I'd lean hard into turning those moments into an in-context answer rather than just another dashboard the creator has to interpret. Does it cluster the confusion points across many viewers so I can fix the top 3, or surface them per session?