Reflection AI is scaling aggressively after its $2B raise, reaching roughly 230–250 staff by mid-2026 3 sources. With massive compute commitments 3 sources, they have the resources to bid for elite talent 2 sources. However, translating total headcount into qualifying hires requires a strict filtering funnel: the resolution criteria demand explicit senior titles (research lead, principal/staff+), which filters out L4/L5 Research Scientists and flat 'Member of Technical Staff' roles unless holding a formal lead title. Public profiles confirm a growing cohort of verifiable senior big-three alumni 44 sources. The wide right tail reflects the potential for aggressive lift-outs from DeepMind London gov.uk. Evaluated alongside the broader scaling narrative, expectations remain largely stable as they are already consistent with the projected capital and compute trajectories. The median of 33 assumes a continued, moderately successful expansion skewed toward verifiable Staff-level deep-learning scientists, tempered by execution risk and extreme compute burn rates, keeping their estimates consistent with their massive capital advantage over smaller peers .