Santa Cruz, California, United States
Hi, my name is Mira! I like to build practical, human-centered systems that help people navigate information, make decisions, and explore opportunities more easily. Most of my work focuses on agentic and retrieval-driven AI, where I design pipelines that combine data engineering, machine learning, and LLMs. I’ve led and contributed to projects in education, career exploration, and decision support, from building community chatbots used by real students to developing full-stack AI applications. One thing that matters most to me in this field is ensuring that AI is used responsibly as a tool to make our lives easier. I do not believe that AI should replace human connection and communication, but should instead encourage and support it. I care about making sure AI systems are designed in ways that promote efficiency, without taking away accuracy and quality.
- Evolution of Community RAG project from Summer '25 - Directed a 6-member engineering team (with 2 PhD mentors) to develop an OpenWebUI-based UCSC community chatbot - Architected Python tool-calling and routing modules that deliver semantic club/event retrieval and contextual recommendations - Devised an LLM-driven auto-invite workflow that interprets dialogue context to place users into more than 5 topic-specific channels, improving relevance and user engagement. - Iteratively refining prompt structure and retrieval logic, decreasing hallucinations and strengthening grounding during testing.
Paper Submission Date ~June 2026 - Co-designed and TA’d UCSC Generative AI in Business Strategy course (140 students); directly mentored 23 students, graded 20+ assignments and provided personalized feedback. - Authored 50% of course lab assignments and 100% of grading rubrics, used to onboard 20+ new student researchers - Conducted 3 interviews and fully transcribed/annotated 28 student transcripts (300+ annotations) (~13 hours of recording) to analyze challenges, workflows, and problem solving of students in the course for research paper data
- Led end-to-end design of a scalable career-guidance platform integrating 200+ professional transcripts, improving accessibility of career insights for local high school and college students through a RAG-powered prototype - Delivered 3+ Figma storyboard/sketch whole-platform iterations, incorporated feedback from 3 PhD students + PI - Managed a team of 10 undergraduates and collaborated with PhD lead and PI through 5 design cycles
- Mentored 20+ students across interdisciplinary reading groups (neural networks, autonomous vehicles) - Provided structured evaluation and feedback on 30+ student research presentations
Tutored 10+ undergraduate students 3-days/week in logic, proofs, and set theory, strengthening CS foundations.
- Applied HCD methods to prototype personal protection product for college students in a team of 3 students from across California - Collected 30+ survey responses, conducted 5+ stakeholder interviews to understand user background - Analyzed 5+ competitor products and identified 20+ potential refinements to meet 2 high priority user needs - Selected as 1 of 9 teams to present at the San Francisco Student Leadership Conference '25
- Full-time internship advancing Tech4Good projects in AI education and Community RAG - Delivered a 6-minute pre-recorded research presentation for the UC Santa Cruz CITRIS Symposium, reaching 50+ industry professionals and researchers - Completed 10+ workshops on professional development, leadership, and project communication
- Led 5+ in-person and online tours of the Baskin Engineering School, presenting programs and resources to prospective students and families. - Represented the engineering department at campus outreach events (e.g., Club Rush), distributing materials and engaging 100+ students about academic and research opportunities. - Strengthened communication and presentation skills by tailoring messages to diverse audiences, from high school visitors to transfer students.