← Home

Divine Keetle-Maloney

Product R&D Leadership — Human-Centered AI Interaction, Evaluation, Grounding & Agentic Systems

San Francisco, CA · divinekmaloney [at] gmail [dot] com · Google Scholar

Profile

Product R&D leader with a track record across Meta Reality Labs, Microsoft Research, and Apple — each time entering at early research and driving it to production. Hired into Apple’s Vision Products Group as the founding Senior Research Scientist on an early R&D initiative, then moved into engineering to lead it through ship — work that contributed directly to Siri’s expressive voices and Apple’s broader conversational AI.

Operating thesis, laid out in my essay “Evaluation Is the Product”: with nondeterministic systems, the only way to genuinely center the human is to measure — every fuzzy interaction goal has to become a concrete, repeatable signal before it can ship. A PhD in Human-Centered Computing is training for defining what “good” means and knowing what feels right from an interaction perspective. Collaborate daily with ML leads across data taxonomy, data generation, model integration, and persona definition — including curating the voice and language datasets that made expressive voices genuinely expressive — and with engineering leads to ensure those integrations land well in product, from application latency to on-device performance. Three patents on AI interactivity, more pending.

Experience

Member of Technical Staff · Thinking Machines

Oct 2026 – Present

Staff Engineer · Apple, Vision Products Group

Sep 2022 – Oct 2026

• Took a 0-to-1 prototype out of research and shipped it as a tentpole Vision Pro feature — first as lead engineer, then in a hybrid PM / lead-engineering role, owning scoping, technical roadmapping, milestone planning, cross-functional coordination, and executive communication alongside the engineering work.
• Contributed perceptual research and evaluation to Siri’s expressive voices (shipped iOS 27) — studies on prosody and expressivity paired with large-scale data curation and quick iterative prototypes built to experience the voices firsthand; that research-plus-prototyping loop directly informed how the assistant’s voice output was tuned.
• Ran head-to-head evaluations of external voice and dialogue models against internal quality bars — findings informed executive-level decisions.
• Defined and named the Research Engineering function within Vision Products Group — a new discipline translating open-ended research questions into automatable evaluation and measurement infrastructure for nondeterministic ML output at consumer scale, including eval standardization and launch readiness.
• Lead through ambiguity in an emerging product space: identify nascent research capabilities, prioritize and take the right risks with actionable data — a Human-Centered Computing background guiding the interaction intuition — and validate through internal prototypes and executive demos, carrying work across the research → prototype → product boundary.

Senior Research Scientist · Apple, Vision Products Group

Jan 2022 – Sep 2022

• Incubated the 0-to-1 project above and led Vision Pro launch research end-to-end: original research design → product development → engineering execution, presenting findings directly to executive leadership. This arc is why the move to engineering — to keep driving the program through ship.
• Submitted three patents during this period on AI-mediated spatial interaction.

Co-Founder · br8krm — venture-backed gaming startup

Jul 2020 – Dec 2020

• Co-founded a real-time social VR wellness platform addressing remote-work isolation and burnout; built and shipped the MVP under tight resource constraints, validating demand that earned an Epic Games MegaGrant and a Y Combinator interview. Owned 0-to-1 product, fundraising, and shipping.

Ethicist (Contract) · Meta Reality Labs

Oct 2021 – Dec 2021

• Built privacy, safety, and ethics frameworks for always-on AI-powered wearables (smart glasses; precursor to Meta Ray-Ban) — direct precedent for coordinating partner data protection and trust requirements within a system architecture.

Research Intern, Microsoft Ada Lovelace Fellow · Microsoft Research AI

May 2019 – Aug 2019

• Researched AI-mediated experiences for empathy and emotional understanding — early work on how conversational AI shapes user affect.

Patents — AI-Mediated Spatial Interaction

• Attention Control in Multi-User Environments — US12548271B2 · Granted Feb 2026. Attention-aware rendering in shared XR spaces.
• Contextualized Visual Search — US12579807B2 · Granted Mar 2026. Intent-aware AR overlays integrated into the physical environment view.
• Virtual Assistant Interactions in a 3D Environment — US20250349070A1 · Published Nov 2025. Spatial AI agent embodiment in XR. Additional patents pending.

Selected Publications

22 peer-reviewed publications at CHI, CSCW, IEEE VR, VRST, and CHI PLAY · 2,518 citations · h-index 16 · i10-index 17. Publishing restricted since joining Apple; all listed work is pre-Apple.

• Freeman, G. & Maloney, D. “Body, Avatar, and Me: The Presentation and Perception of Self in Social Virtual Reality.” PACM HCI (CSCW), 2021.
• Maloney, D., Freeman, G., & Wohn, D.Y. “Talking without a Voice: Understanding Non-verbal Communication in Social Virtual Reality.” PACM HCI (CSCW), 2020.
• Freeman, G., Maloney, D., Acena, D., & Barwulor, C. “(Re)discovering the Physical Body Online.” CHI, 2022. ★ Best Paper Honorable Mention.

Education

Ph.D., Human-Centered Computing — Clemson University, 2021
Dissertation: “A Youthful Metaverse” — safe, equitable, and emotionally fulfilling social VR for younger users. Microsoft Ada Lovelace PhD Fellowship; $320K+ in fellowship funding; top 1% of PhD students in CS-related fields.

B.S., Computer Science & Spanish — Sewanee: The University of the South, 2017

Skills

Product: 0-to-1 strategy and validation, MVP scoping, research-to-product translation, program design, roadmaps and milestone planning, executive communication, vendor and acquisition evaluation, patent strategy

ML/LLM systems: Model output evaluation and quality-bar setting, grounding, citation-structure and content-integrity assessment, agentic interaction design, nondeterminism instrumentation, human-computer interaction and interaction models, speech and voice models (Siri expressive voices, above)

Technical: Python (production), Swift (native), TypeScript, SQL; evaluation and measurement infrastructure for nondeterministic AI systems; AR/VR development including Apple Vision Pro