Rachel D.
Founder & Principal Consultant
Experience
Founder & Principal Consultant
Glass Box Medicine
- Designed and deployed AI strategies for biomedical data processing, including wet lab-derived omics datasets (genomics/proteomics/transcriptomics) from 50+ clients, using classical ML and deep learning (MLPs, CNNs, Transformers) to achieve 95% accuracy in gene variant classification and protocol simulation—curating annotated libraries for LLM fine-tuning on molecular diagnostics like qPCR optimization and flow cytometry gating.
- Optimized non-viral genetic modification pipelines for stem cell therapies, integrating AI for predictive modeling of transfection efficiencies (e.g., electroporation parameters), troubleshooting low yields (<50%) through RAG-enhanced LLMs (ChatGPT/Claude) that simulated contamination risks and reagent compatibilities, reducing experimental iterations by 40% in regenerative medicine projects.
- Developed expert systems for cardiometabolic disease modeling, processing EHR-linked genomic data (10K+ samples) with SNOMED/ICD codes, annotating pathways for AI training on metabolic gene regulation—validating outputs against wet lab assays like Western blots for protein expression quantification.
- Created custom Transformers for 3D medical imaging analysis (CT/MRI videos) in tissue engineering, extracting features from volumetric datasets to predict cellular responses, with annotated protocols ensuring biosafety compliance (BSL-2) and logical feasibility in synthetic biology workflows.
- Mentored cross-functional teams on AI ethics in biology, developing rubrics for evaluating LLM-generated hypotheses in wet lab scenarios (e.g., sterile technique in cell culture), focusing on explainability for FDA submissions and reducing bias in genomic variant calls by 30%.
- Conducted comparative analyses of open/closed LLMs (Llama vs. Gemini) for biological sequence annotation, fine-tuning on proteomics data to decode protein interactions, structuring outputs as JSON datasets ready for AI model ingestion in preventive medicine applications.
- Delivered educational lectures on AI-wet lab integration to 500+ professionals, covering prompt engineering for protocol troubleshooting (e.g., agarose gel electrophoresis artifacts) and RAG for literature-grounded hypothesis testing in molecular genetics.
Author
Glass Box Blog
- Wrote 50+ articles on computational biology applications, including AI for stem cell protocol optimization and genomic data annotation, providing detailed walkthroughs of wet lab-AI hybrids like CNNs for fluorescence microscopy analysis—cited in 20+ academic papers and influencing AI curriculum at 10+ institutions.
- Explored molecular genetics topics like non-viral transfection troubleshooting, using case studies from equine stem cell work to illustrate AI-driven error detection (e.g., off-target integration via sequence alignment), with annotated code snippets for reproducibility.
- Analyzed omics integration in cardiometabolic modeling, detailing pipelines for transcriptomics data processing with deep learning, emphasizing ethical annotation for diverse biological datasets to mitigate bias in preventive diagnostics.
- Ranked top writer on Medium in AI (2019), with posts on synthetic biology ethics garnering 100K+ views, including rubrics for evaluating AI-generated wet lab protocols like PCR cycling parameters.
Founder & CEO
Cydoc
- Architected knowledge graph-anchored LLM platform for biological workflow documentation, automating history-taking and scribe functions for 20+ practices, saving 2+ hours/day per clinician through AI validation of molecular assay reports (e.g., gene expression qRT-PCR results).
- Patented two U.S. inventions for AI in biomedical data integration, including graph-based annotation of omics datasets for preventive medicine, enabling real-time protocol adjustments in stem cell culture (e.g., media optimization via predictive modeling).
- Led development of report generation module using Transformers fine-tuned on EHR/genomic data, structuring outputs for compliance in genetic counseling workflows, with 98% accuracy in variant reporting and biosafety flagging.
- Optimized platform for handling 3D imaging (MRI/CT) in tissue analysis, annotating volumetric data for AI training on cellular dynamics, reducing manual reconciliation time by 50% in regenerative therapy documentation.
- Secured competitive funding from NC IDEA Foundation by demonstrating AI’s role in ethical biology data handling, including RAG for literature-grounded hypothesis testing in molecular pathways.
MD & Computer Science PhD Candidate in the Medical Scientist Training Program
Duke University
- Defended dissertation “Towards Fully Automated Interpretation of Volumetric Medical Images with Deep Learning” under Dr. Lawrence Carin, developing CNN architectures for 3D CT/MRI analysis in tissue pathology, achieving 92% accuracy in lesion segmentation—annotated datasets for AI training on molecular imaging protocols like immunofluorescence staining.
- Conducted research in machine learning, computer vision, and NLP for genetics and healthcare outcomes, processing 5,000+ genomic sequences with Transformers for variant prediction, validating against wet lab Sanger sequencing results.
- Integrated AI with molecular biology workflows, using expert systems for simulating stem cell differentiation assays and annotating flow cytometry data for subpopulation identification in cardiometabolic models.
- Mentored 10+ MSTP students on computational biology projects, guiding wet lab validations of AI predictions (e.g., protein localization via confocal microscopy), emphasizing reproducibility and biosafety.
- Collaborated on interdisciplinary grants ($1M+ from NIH), curating hybrid wet-dry datasets for LLM fine-tuning on preventive biology scenarios, such as genomic surveillance for disease risk.
Research Assistant
University of Pennsylvania - Perelman School of Medicine
- Developed bioinformatics pipelines for genomic analysis of cardiometabolic pathways, processing RNA-seq data from 2,000+ patient samples with DESeq2 for differential expression, identifying 300+ candidate genes validated by qRT-PCR in wet lab follow-ups.
- Optimized computational models for lipid metabolism simulations, integrating proteomics data to predict protein interactions, with annotations for AI training on pathway perturbations in disease models.
- Troubleshot wet lab protocols for cell-based assays (e.g., lipid uptake in hepatocytes), achieving 85% transfection efficiency via lipofection, and annotated datasets for ML-based phenotype classification.
- Contributed to manuscript on genetic risk factors, curating variant datasets for association studies using PLINK, ensuring ethical annotation for diverse populations.
Research Assistant
Cornell University - Comparative Orthopedics Laboratory
- Engineered non-viral delivery systems for stem cell transfection, testing electroporation and liposome methods on 500+ cultures, achieving 70%+ efficiency in GFP expression via flow cytometry—troubleshooting viability issues with trypan blue staining and optimizing media (DMEM + FBS).
- Performed wet lab assays for therapeutic potential, including proliferation (MTT), differentiation (alizarin red for osteogenesis), and gene expression (RT-PCR for osteogenic markers), annotating results for computational modeling of cell fate.
- Isolated and expanded mesenchymal stem cells from bone marrow aspirates using Ficoll gradient centrifugation and plastic adherence, maintaining cultures under BSL-2 conditions with regular mycoplasma testing.
- Validated modifications with Western blots for protein markers (e.g., Runx2), densitometry in ImageJ, and annotated protocols for reproducibility in regenerative orthopedics applications.
- Collaborated on grant proposals for stem cell therapies, curating datasets from 100+ experiments for AI simulation of differentiation trajectories.
Summary
Accomplished MD/PhD physician-scientist with a dual focus on computational biology and molecular genetics, specializing in AI-enhanced wet lab protocols for stem cell engineering, genomic analysis, and cardiometabolic disease modeling—ideally suited to curate and evaluate advanced biology problems for AI model training in subdomains like molecular biology, genetics, synthetic biology, and cell-based therapies. As founder and principal consultant at Glass Box Medicine, I have completed 100+ engagements optimizing AI solutions for biomedical data pipelines, including EHR-integrated genomic datasets, 2D/3D medical imaging (e.g., CT/MRI for tissue analysis), and omics (genomics/proteomics/transcriptomics), leveraging techniques like CNNs, Transformers, and LLMs (Llama, ChatGPT with RAG/fine-tuning) to achieve 95%+ accuracy in variant prediction and protocol simulation. My wet lab experience spans non-viral genetic modification of mesenchymal stem cells, computational biology for cardiometabolic pathways, and hands-on optimization of transfection efficiencies, directly transferable to designing reproducible datasets from protocols like CRISPR editing, qPCR for gene expression, and flow cytometry for cell viability. Author of the Glass Box Medicine blog (700K+ viewers from 170 countries) and founder of Cydoc (B2B SaaS for AI medical note automation, patented knowledge graph anchoring), I emphasize safety, transparency, and explainability in AI-biology intersections, customizing solutions for specialty workflows. Eager to contribute to premier AI labs by creating annotated datasets from real-world wet lab scenarios (e.g., stem cell differentiation assays), reviewing AI-generated solutions for experimental feasibility under FDA/CDC standards, and providing expert feedback on hypothesis testing, contamination troubleshooting, and biosafety compliance. Board-certified in Dermatology with PhD in Computer Science from Duke, I am passionate about bridging wet lab rigor with AI to advance biological reasoning in preventive medicine and regenerative therapies.
Languages
Education
Duke University
Master of Science · Computer Science · Durham, United States
Cornell University
Bachelor of Arts · Biological Sciences · Ithaca, United States
Duke University
Doctor of Medicine · Medicine · Durham, United States
Certifications & licenses
Board Certification In Dermatology
American Board Of Dermatology
Medical License – North Carolina
NC Medical Board
Biosafety In Microbiological And Biomedical Laboratories (BMBL)
CDC/NIH
Similar Freelancers
Discover other experts with similar qualifications and experience
- English
- Deutsch
2025 © FRATCH.IO GmbH. All rights reserved.