Revenue Forecasting & Strategic Modeling
I build B2B revenue forecasting and strategic models that help surface demand patterns, sharpen planning, and connect machine learning work to concrete growth decisions.
ML Scientist at OpenAI
Hey, I'm Jackie
I build revenue intelligence and grounded AI systems
I’m an ML Scientist at OpenAI working on B2B revenue forecasting and strategic modeling to better understand business demand, inform growth decisions, and turn noisy signals into decision-ready insight. Alongside that work, I stay deeply invested in AI systems design, evaluation, retrieval, and open-source tools that help people learn faster, read research more effectively, and work more intelligently with data.
Previously, I worked across forecasting, experimentation, recommendation systems, and AI product development at Microsoft, including context engineering and evaluation for Copilot Notebooks. I also collaborate across health AI and aging-related work, and build self-driven open-source tools such as codex-skills, ArxivSummary, and schema-lineage tooling.
More about my backgroundWhat I Work On
Building impact across work and AI side projects.
My core role is business-critical machine learning. My side work stays broad, self-driven, and deeply practical.
Applied AI Systems & Evaluation
I’ve worked end to end on AI product development, including context engineering, prompt design, tool usage, and evaluation for grounded systems such as Microsoft Copilot Notebooks.
Open Source, Research Tools, and Data Workflows
I build open-source AI tooling and research-facing systems that help people synthesize papers, mine data more effectively, and turn technical curiosity into useful software.
Recent News
Recent milestones across work, research, and open source.
A quick snapshot of the latest updates. The full archive lives on the news page.
Link
March 2026
January 2026
December 2025
View all news
Talk at UCSF Biostatistics Lab
Had a discussion with the UCSF Biostatistics lab on how research training carries into AI-native work, including problem formulation, validation, and the role of human judgment in real data and modeling workflows.
Link March 2026Codex Skills Series Launched
Launched a new codex-skills series with reusable, installable workflows for data mining, modeling, AI, and research workflows, making capabilities easier to reuse across projects with less setup overhead.
Joined OpenAI
Started a new chapter as an ML Scientist at OpenAI, working on B2B revenue forecasting and strategic modeling.
NeurIPS 2025
Attended NeurIPS 2025 and published notes on research trends, evaluation, and practical LLM takeaways.
Contact
Interested in AI products, modeling, or research tooling?
I’m especially interested in work that connects technical rigor to real outcomes, whether that means better revenue insight, stronger AI product behavior, or more useful open-source systems.