{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/6b7aafc93d4a4a73a231ecae49d60ce8\" frameborder=\"0\" width=\"1280\" height=\"960\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":960,"width":1280,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":960,"thumbnail_width":1280,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/6b7aafc93d4a4a73a231ecae49d60ce8-c28ea1d00c7b1dd3.gif","duration":300.201,"title":"CohortFilter AI for Startup Accelerator Screening","description":"This Loom demonstrates Forward Filter AI, an intelligent screening tool that ranks startup applications for accelerator and VC teams. The tool takes a CSV upload of 100 to 1,000 applications, uses a configurable chat interface for deal-breakers and scoring weights, and then produces ranked shortlists with scores, reasoning, and “reflex” in minutes. Using the MI 300X with VLM on ROCM and full precision, it can score 50 applications in about 30 seconds, and thousand-application runs are expected in under five minutes. Results can be exported as a PDF or CSV, and the author emphasizes customization and data sovereignty with no data leaving the environment."}