Amy Ji

Quantitative Biology and Bioinformatics graduate student bridging molecular biology and computational genomics, with experience spanning next-generation sequencing analysis, single-cell and machine learning workflows. I build bioinformatics pipelines in Go, Python, and R, and I am drawn to agentic AI workflows as a way to automate and accelerate the tedious parts of daily research. My wet lab foundation includes PCR, gel electrophoresis, mammalian cell culture, and sample preparation, all grounded in careful, SOP-driven laboratory practice.

I am seeking a job or research position in bioinformatics or a related interdisciplinary field in biomedicine.

Graduate Student Researcher - Mellon Institute of Science, Cai Immunology Lab - Pittsburgh, PA

(2026-01)

Support T-cell immunology studies through maintaining mammalian cell culture, practicing sterile technique, and preparing experimental samples.

• Adapting a pre-trained, U-NET++-based deep learning model (FAST) with a TIRF dataset to quantify the number and length of T-cell microvilli to support lab research.
• Automated repetitive image-processing tasks such as channel filtering, splitting, and adjustment using ImageJ Macro Language (.ijm), enabling analysis of hundreds of TIRF microscope results within minutes.
• Developed and tested a safe and efficient lentiviral packaging system for murine tumor cell line transduction.

AI Engineering Intern, Marketing Intelligence - EnquBio Inc - San Diego, CA (hybrid)

(2026-05 - 2026-08)

• Designed and built an AI agent backed by a modular Go engine that automated pharmaceutical intelligence gathering for a contract research organization, replacing manual research across 3 sources (ClinicalTrials.gov, PubMed, EDGAR) with structured, queryable signals and cutting analyst research time by ~70%, from hours to under 10 minutes per query.
• Deployed a fully local LLM extraction (Qwen 3.6 via Ollama) and embedding (mxbai-embed-large) stack on an NVIDIA DGX, processing ~200 documents/day while keeping proprietary data and conversations on premises and eliminating cloud API costs.
• Built a self-improving personalization layer that scores each biomedical signal per user across 8+ weighted features and updates preferences from analyst feedback stored in PostgreSQL, lifting signal relevance by ~40% over the baseline ranking.
• Exposed the engine as a per-user, identity-secured MCP tool server offering 13 tools, enabling an AI agent to query tracked entities, retrieve personalized signals, and capture feedback through natural language across 10+ daily active users.

Computing and AI in R&D, Intern - Qilu Pharmaceutical - Shanghai, China

(2025-06 - 2025-08)

• Automated extraction of biomedical and pharmaceutical information from unstructured documents using OCR (pytesseract) and LLM-assisted parsing (DeepSeek-R1), reducing manual review time by approximately 80%.
• Structured key R&D fields such as protein targets, disease indications, mechanisms of action, and expiration-related metadata into analysis-ready datasets for downstream research use.
• Built a pipeline to systematically store and organize patent filings and drug-development records, enabling efficient cross-referencing and retrieval across therapeutic programs.
• Improved data-processing quality through iterative validation with cross-functional stakeholders in research and development teams.

Computational Research Assistant - Boston University Chobanian & Avedisian School of Medicine, Dept. of Nephrology - Boston, MA

(2023-09 - 2025-05)

• Contributed to a deep learning image-segmentation model (92% accuracy) for transmission electron microscopy kidney images, automating measurement of glomerular basement membrane and podocyte foot process structures.
• Curated and labeled 400+ training images with domain-specific annotations for model validation.
• Performed statistical analysis and data visualization in Python and R to evaluate model performance; results contributed to a manuscript currently under peer review.
• Independently operated a Hitachi Scanning Electron Microscope for tissue sample imaging.
• Optimized the sample preparation protocol by replacing osmium-based fixation with a safer potassium-based alternative, reducing cost and hazard exposure.

Learning Assistant, Systems Physiology - Boston University - Boston, MA

(2024-09 - 2025-05)

• Supported over 40 students per semester, improving lab software proficiency, experimental design, and scientific writing.
• Increased student performance on lab assessments by 15%, with 70% reporting improved understanding.
• Collaborated with faculty in weekly meetings to refine teaching strategies and course delivery.

Research Intern - Evonik Specialty Chemicals - Shanghai, China

(2023-06 - 2023-08)

• Synthesized over 30 chemical samples for skincare product development, tested viscosity, and monitored stability across temperature conditions.
• Translated technical documents and supported international client communications.

Master of Science (MS) - Quantitative Biology and Bioinformatics - Carnegie Mellon University (2025-09 - 2026-12)

Bachelor of Arts (BA) - Biochemistry and Molecular Biology - Boston University (2021-09 - 2025-05)