Description and Requirements
. Lead technical research and experimentation in User World Models, predictive AI, knowledge graphs, and data representations for Personal AI.
. Formulate hypotheses, reproduce relevant research, build baselines and measure improvements in context understanding, prediction quality, grounding and reliability.
. Develop privacy-aware representations of user activities, relationships, preferences, goals, temporal context and cross-device state account for uncertainty, consent and changing information.
. Explore graph-based, semantic and temporal memory, entity resolution, data fusion, retrieval and reasoning across permitted application and device data.
. Investigate predictive models that anticipate user needs or next actions, with clear confidence thresholds, evaluation and user control.
. Partner with prototype and systems engineers to validate ideas in realistic Qira experiences on Lenovo devices, including proactive and cross-device scenarios.
. Contribute to patents, technical disclosures and, where appropriate, external research publications.
Basic Qualifications
. Bachelor's or Master's degree in Computer Science, AI, ML or a related technical field PhD or equivalent research experience is a plus.
. 5+ years of relevant AI/ML experience, with strong evidence of hands-on model development or research engineering.
. Deep experience with modern LLM/SLM or multimodal AI systems and frameworks.
. Experience in one or more of user modeling, personalization, predictive modeling, knowledge graphs, information retrieval, LLMs or contextual reasoning.
. Strong Python and software engineering skills comfortable taking ideas from experiment to working code.
. Ability to reason about model quality, latency, cost, reliability, privacy and real-world constraints.
Preferred Qualifications
. Experience with temporal/heterogeneous knowledge graphs, graph ML, entity linking, probabilistic reasoning or causal/predictive modeling.
. Experience with multimodal context, memory systems, privacy-preserving learning or cross-device data integration.
. Experience with multimodal foundation models, VLMs, speech/audio or vision-language systems.
. Publication, patent or open-source contributions in AI/ML.
Success Measures
. Deliver reproducible research experiments and working prototypes for User World Models, Predictive AI and Knowledge Graph/Data priorities.
. Demonstrate measurable improvements in contextual understanding, prediction usefulness, grounding, reliability or privacy-aware personalization.
. Demonstrate a clear path from promising experiments into advanced development or product adoption.
Key Partnerships
Works closely with the Lead Scientist/Research leadership, AI Systems, Prototype Engineering and global Qira Product/Engineering teams.
