Perfectly

The problem: Traditional recruiting agencies are slow, manually bottlenecked, and lose signal between candidates, recruiters, and hiring managers. The founders describe sitting through 800+ agency‑sourced interviews at TikTok, observing that pipelines moved slowly and candidates were often mismatched. From Perfectly’s FAQ and positioning, agencies commonly deliver just a handful of candidates per week, struggle with deep domain understanding, and deprioritize hard roles.

Perfectly’s solution: Perfectly presents itself as the first AI‑native, end‑to‑end recruiting agency and operating system. It conducts a deep intake to capture nuanced preferences, then uses an AI system to source at scale, automate outreach to thousands of candidates, conduct a reusable AI screen, and feed interview‑ready candidates directly to the hiring manager’s portal (and even into Slack). The system is designed with “infinite memory and context,” so feedback from each role improves future matching. Claims include delivering 5‑10 interview‑ready candidates within 24 hours, doubling interview pass rates vs. traditional agencies, and filling roles 4x faster—often filling the interviewing schedule the same day as intake.

Commercial model and differentiation: Perfectly only charges when a role is filled and quotes a 15‑25% fee, aiming to undercut traditional agencies by about 50%. This aligns with the founders’ thesis: if AI eliminates the manual bottlenecks and reduces error, it can justify faster, better, and cheaper outcomes simultaneously. Testimonials reinforce impact (e.g., a Series A client that fired other agencies; moving off Paraform and canceling Juicebox due to Perfectly’s velocity).

Why it matters: For startups with urgent, hard‑to‑fill technical roles, shrinking time‑to‑fill from months to weeks (or days to pipeline) while improving pass‑through quality is a material competitive advantage. The approach also compounds: continuous feedback, structured data capture, and reusable screens increase accuracy over time, turning recruiting from a sporadic, human‑memory constrained process into a scalable, data‑driven system.

Market context and trends: The global staffing market is massive (~$650B 2022) but traditionally weighted toward temporary staffing; the permanent/direct placement segment is still tens of billions and ripe for modernization via AI and automation. U.S. staffing sales remain very large despite cyclicality ($142.4B temp/contract in 2023). Adjacent models like RPO are growing quickly (16.1% CAGR to 2030), and AI‑first recruiting software categories are scaling, suggesting budget reallocation toward AI‑native solutions.

Competitive landscape by model:

• Marketplaces/agency networks: Paraform matches companies with elite recruiters and handles business development so recruiters can focus on placements—essentially a marketplace that aggregates agency capacity.
• AI recruiting software platforms: Juicebox (AI recruiting platform with Search/CRM/Agents), hireEZ (agentic AI from application‑to‑interview: screening, fraud detection, AI phone screens, scheduling), Gem (AI‑first all‑in‑one including ATS/CRM/sourcing/scheduling/analytics), Loxo (AI‑powered Talent Intelligence Platform with ATS/CRM/source/outreach), Eightfold (Agentic Talent OS with talent intelligence across TA and TM). Fetcher blends AI with an expert human team for sourcing.
• Talent marketplaces: Hired, Triplebyte, Turing (not analyzed in detail) focus on candidate pools and matching but differ in economics and workflows vs. a full‑service agency.

Perfectly’s positioning and advantages: Perfectly occupies a hybrid niche: it is an agency in commercial model (fee on placement) but AI‑native in operations, claiming an end‑to‑end system that matches, outreaches at scale, screens with AI, and routes candidates straight to hiring managers. This allows Perfectly to compete against traditional agencies on speed/quality/price and against software‑only platforms by delivering outcomes (hires) rather than tools. The founders’ recommender‑system pedigree from TikTok/Meta strengthens the tech narrative and fits the claim of “infinite memory and context” with improving accuracy over time. Pricing that undercuts agencies by ~50% (while quoting 15‑25%) is explicitly designed to break incumbent tradeoffs.

Risks and considerations: The market is crowded with AI recruiting software and marketplaces; switching costs and trust are key. Perfectly’s strong performance claims (2x interview pass rate; 4x faster fills) need continued proof with references, logos, and quantified case studies. Macroeconomic cycles can compress hiring volumes, but AI‑native efficiencies may gain share during downturns as teams seek more output per dollar.

Founders: Perfectly was founded in 2025 by Victor Luo and Zhuang (Gary) Luo and is part of Y Combinator’s Winter 2026 batch in San Francisco. YC’s profile and launch post explicitly list both founders and describe the company as the first end‑to‑end AI recruiting agency. Their shared background is ML and large‑scale recommender systems, which directly informs Perfectly’s algorithmic approach to talent matching.

Victor Luo: YC lists Victor as a founder and “2x founder | ex‑ML Scientist at TikTok.” His public LinkedIn header indicates Perfectly (YC W26) and University of Virginia as his school affiliation. The YC launch post further explains the founders’ perspective on agency recruiting after sitting through 800+ agency‑sourced interviews at TikTok, which motivated them to rebuild the process using modern AI methods. These details establish Victor’s prior experience working on advanced recommendation systems and his exposure to the pain points of traditional recruiting from the hiring team’s side.

Zhuang (Gary) Luo: YC lists Gary as “ex‑MLE at Meta and TikTok” and co‑founder of Perfectly. A professional profile on The Org adds detail: he’s a skilled ML engineer with roles focused on recommendation optimization (TikTok/ByteDance), data‑driven growth at Soul App, and broader ML strategy; it also notes a Bachelor of Engineering from Zhejiang University and Master’s studies at Worcester Polytechnic Institute. This path is consistent with building and deploying production‑grade ML models at scale—highly relevant to building an AI‑native recruiting engine.

Team DNA and credibility: Perfectly’s site repeatedly emphasizes the team’s backgrounds as ML scientists from TikTok and Meta and frames the product as applying recommender‑system science to hiring. This connects the founders’ technical histories directly to Perfectly’s differentiation: modeling candidate‑job fit, predicting interview/job performance, and continuously improving via feedback (“infinite memory and context”).