Anushay Anjum

Summer 2026

• Served across product management, data science, and engineering for the Surface MCP Server, owning adoption and reliability from telemetry instrumentation and product requirements through analysis, dashboards, and stakeholder execution.
• Built a Kusto (KQL) and Power BI adoption-and-reliability dashboard tracking active devices, launch-to-tool conversion, completion and success rates, P50/P95 latency, failures, and app-version adoption; added in-app missions to improve onboarding, activation, and self-host engagement.
• Audited and repaired telemetry across battery, device, keyboard, and pen tools, resolving naming mismatches, missing parameters, and misleading success signals while defining canonical events and tool-specific metrics with engineers and PMs.

Summer 2025

• Replaced a 35-year-old Surface battery-life model with a forecasting system built in Python and Kusto over 10M+ telemetry rows; engineered 75+ features and a new time-series calculation method, cutting error from 22% to under 10% MAPE.
• Tuned 500+ LightGBM models with time-series cross-validation, reducing the RMSE gap by 30% and reaching 75%+ negative-bias reliability.
• Productionized the pipelines through Microsoft MLOps CI/CD with automated retraining, monitoring, rollback-safe deployment, and a post-production buffer calibration layer.
• Mentored underrepresented students through the Black@Microsoft ERG, teaching data-science and machine-learning fundamentals.

Summer 2024

• Analyzed single-cell RNA sequencing, lineage-tracing, and CNV data with Python (Pandas, NumPy) and R (Seurat).
• Applied supervised, unsupervised, and deep-learning methods to uncover novel insights.
• Used feature engineering and dimensionality reduction including PCA and t-SNE, improving model performance by 25%.

Spring / Summer 2024

• Worked in spatial computing and XR with Unity3D and C#, integrating PyTorch deep neural networks into customer-centered AR, VR, and MR experiences.
• Developed and presented a VR healthcare solution using Oculus SDK and REST APIs to senior engineering executives; the proposal created immersive virtual clinics, reduced geographical barriers by 60%, and integrated with EHR and telehealth platforms.

Spring 2024

• Conducted data-science and neuroanalytics research under Dr. Pascal Wallisch using data from more than 1,200 participants.
• Studied trauma-aggression reciprocity with VADER sentiment analysis and machine learning across more than 500 hours of college-student audio recordings.

Fall 2023

• Built a scikit-learn and Pandas model from cognitive data and food-choice histories for 500 geriatric rhesus macaque monkeys.
• Forecasted food choices with 80% accuracy, supporting research into primate nutritional decision-making.

Fall 2023

• Implemented a Python Random Forest classifier over 10,000+ cardiovascular patient records, improving the speed and precision of heart-disease diagnosis by 30%.

Summer 2023

• Used R (dplyr, ggplot2), PostgreSQL, Python, Pandas, and Matplotlib in research on neurological conditions and neurotechnology.
• Contributed to experimental design, collection, and analysis across neurorobotics, prosthetics, BCIs, EEG, ECoG, and EMG, processing datasets larger than 500GB.

Leadership

• Worked with a team of five to mentor more than 50 young girls in technology.
• Facilitated hybrid in-person and virtual sessions to support different learning needs and maximize engagement.

Leadership

• Mentored more than 100 high-school students through technology research projects, from inquiry and analysis to presentation.

Teaching

• Taught 50 underrepresented Pakistani middle- and high-school students remotely over Zoom.
• Provided academic support and mentorship that helped students improve their grades by 30%.