Traditional human coaching remains inaccessible to most early-career professionals due to scalability and cost constraints, while recent evidence suggests AI coaching platforms can deliver meaningful developmental outcomes, yet rigorous experimental validation remains limited. This randomized controlled trial will evaluate the effectiveness of an AI-powered coaching platform on workplace performance, well-being, and career development outcomes among early-career professionals across three regions: Europe (n=300), United States (n=300), and India (n=300). Participants will be randomly assigned to either: (1) AI coaching intervention group receiving 12 weeks of personalized coaching via the Sherlock AI platform, or (2) waitlist control group. The AI coaching system integrates behavioral science frameworks with large language models, providing psychological safety-focused dialogue, adaptive inquiry, and personalized developmental pathways. Primary outcomes include work engagement (Utrecht Work Engagement Scale), psychological well-being, PERMA+4, and self-reported career development confidence measured at baseline, 4 weeks, 8 weeks, 12 weeks, and 3-month follow-up. Secondary outcomes encompass 360-degree feedback on communication effectiveness, emotional regulation, and leadership behaviors; organizational commitment; turnover intention; and platform engagement metrics including coaching session frequency and duration. Based on preliminary pilot data showing 92% coaching effectiveness ratings and significant engagement (13,700+ minutes across 900 users), we hypothesize significant improvements in primary outcomes for the intervention group, with potential cultural variations in response patterns. This study will provide the first large-scale experimental evidence for AI coaching effectiveness across diverse cultural contexts, informing evidence-based implementation of scalable developmental interventions for the global workforce.