Cognify offers AI‑driven MCAT agent

Polsia promoted a new MCAT study product called Cognify on May 22, presenting it as an AI agent rather than a static question bank or study calendar. The post said the tool monitors a student’s knowledge gaps, generates personalized practice and pushes a daily study schedule. (x.com) The pitch places Cognify in a crowded test-prep market that already includes banks of practice questions, fixed study plans and full-length exams. The launch also arrives as universities and professional schools are writing stricter rules for when students can use generative AI. UC Berkeley School of Law said in a policy effective Summer 2026 that AI is barred for conceptualizing, outlining, drafting, revising, translating or editing work submitted for credit, and for any use in exams. (law.berkeley.edu) ### How is Cognify being described? The May 22 Polsia post described Cognify as an AI agent for MCAT preparation that watches for weak areas, creates targeted practice and assigns daily work. That framing differs from a conventional prep product that gives every user the same sequence of chapters or question sets. (x.com) The post also said the builder had a 99th-percentile MCAT background and designed the product to adapt to each user’s weaker topics. That claim, as presented in the post, is part of the product’s sales pitch and was not independently substantiated in the materials reviewed. (law.berkeley.edu) ### What problem is the tool trying to solve? MCAT preparation products have long emphasized error review, weak-area tracking and schedule discipline. Recent social posts from test-prep accounts and medical trainees have pushed similar themes: review wrong answers closely, identify recurring gaps and keep a consistent study routine. (x.com) Cognify’s stated proposition is to automate that process. Instead of asking a student to manually log mistakes and rebuild a plan, the tool says it can detect where the student is falling behind and produce the next set of tasks. (x.com) ### How does this fit into the broader AI-in-education debate? UC Berkeley Law’s new AI policy offers a concrete example of the institutional mood around AI-assisted academic work. The policy says “thinking remains the sine qua non” of legal education and bars AI help on most work submitted for credit, while allowing limited deviations when instructors intentionally teach AI fluency. (x.com) That distinction matters for study tools such as Cognify. A product that helps students quiz themselves, summarize their own notes or organize practice may fit within many schools’ rules, but a product that crosses into drafting or completing graded work could conflict with campus policy, depending on the institution and course. (x.com) That is an inference from the Berkeley policy’s language, not a statement by Cognify or Polsia. ### What remains unverified? The available public material reviewed for this story did not establish a standalone official Cognify product page tied to the May 22 MCAT post, nor did it provide pricing, launch timing, institutional partnerships or usage figures. (law.berkeley.edu) Searches for “Cognify” returned several unrelated companies and software products, underscoring that the MCAT tool’s public footprint is still limited in the sources reviewed. Polsia’s own website describes the company more broadly as an autonomous AI system that plans, codes and markets companies, but the site material surfaced in search did not itself provide detailed documentation for the MCAT tool. (law.berkeley.edu) ### What should readers watch next? The next concrete markers are a public product page, pricing or access terms, and any disclosure from Polsia or the creator about how Cognify uses student data and what guardrails it applies. School-level AI rules are also moving: Berkeley Law’s policy takes effect in Summer 2026, and similar policies elsewhere could shape how students use tools like Cognify in coursework and exam preparation. (getcognify.ai) (law.berkeley.edu) (polsia.com)