Illustrated portrait of Julie Hohenberg

Julie Hohenberg

I’m not shy. I’ve done cold calls, pitched to VCs, and presented to an audience of college students about the female anatomy.

I love learning cool things: my latest rabbit hole.

I am human-centered with everything I do.

Scroll to learn more

Experience

Download resume ↓
Chewy AI FellowJuly 2026 to Present Boston, MA
  • Conduct internal discovery to identify high-impact problems and disruptive solutions to present to the VP of AI.
  • Owned development of an agentic clinical-audit framework for Chewy Health with clinician-in-the-loop validation and integrated clinical search, projected to save over 3,700 clinical hours annually.
Revion Co-Founder & Venture LeadNovember 2025 to May 2026 Boston, MA
  • Cold-called clinics and secured data-sharing partnerships with two high-volume breast augmentation practices.
  • Built a computer vision model to generate post-surgery breast augmentation visualizations from pre-surgery images.
  • Won first place at the Boston College Elevator Pitch Competition, securing $1,500 in funding from SSC Venture Partners.
View project ↓
Ranger Ultrasound-AI Lab Research AssistantMay 2024 to May 2026 Chestnut Hill, MA
  • Developed a deep learning U-Net model to predict body composition from ultrasound images, reaching over 80% accuracy across multi-site hospital data.
  • Led model interpretability using Grad-CAM, serving as a technical-clinical liaison to align technical work with clinical requirements.
  • Worked with hospitals in Minnesota, Boston, and Ethiopia to publish multiple peer-reviewed papers.
Visit Ranger Lab ↗See research project ↓
Dell Technologies Data Science InternJune 2025 to August 2025 Hopkinton, MA
  • Developed a sentiment analysis model trained on historical escalation data to identify high-risk service cases, achieving 75% accuracy in flagging high-risk cases.
Calico Care AI ConsultantJune 2025 to August 2025 Boston, MA
  • Developed a nurse-in-the-loop AI triage system to analyze older adults' biometric data and deliver personalized, supportive feedback.
Prud'hommeaux Computational Linguistics Lab Research AssistantJune 2023 to May 2024 Chestnut Hill, MA
  • Developed an NLP model to rate utterances on a scale of 1 to 3 for politeness, uncertainty, and information content, informing future computational linguistics research.
  • Co-authored a poster accepted to the International Society for Autism Research 2025 Annual Meeting.
See research project ↓
Some research projects I'm really passionate about

Ultrasound-AI Research

Built a deep-learning system that predicts body composition directly from ultrasound images, then used Grad-CAM to make the model’s reasoning visible.

My roleI developed the model and led the interpretability work. Using Grad-CAM, I checked which parts of each ultrasound image influenced its predictions.

CollaborationI worked with technical and clinical stakeholders, including hospital teams in Boston, Minnesota, and Ethiopia. I led communication between the groups as we reviewed results and prepared multiple peer-reviewed papers.

Grad-CAM visualizations showing the ultrasound image regions that most influenced the model's body-composition predictions.
Grad-CAM heatmaps reveal the image regions most influential to each prediction, where FM is fat mass and FFM is fat-free mass.

Natural Language Processing

How Do Virtual Communication Settings Impact Pragmatic Language in Autistic Adults?

I co-authored research on how in-person and video communication settings shape pragmatic language, including politeness, uncertainty, and informativeness.

Presented as a poster at INSAR 2025 ↗

I started by rating utterances by hand for politeness, uncertainty, and information content. I saw an opportunity to scale that work, so I used those ratings as ground truth to fine-tune BERT. The model reached over 85% accuracy and gave the lab a way to annotate more language for future research.

Research poster titled How Do Virtual Communication Settings Impact Pragmatic Language in Autistic Adults? Open poster in a new tab ↗

Revion

Julie presenting Revion with her team at Boston College's Elevator Pitch Competition.
Julie presenting Revion with a microphone.
Julie and a Revion teammate holding first-place competition checks.

Making breast augmentation consultations less uncertain.

I co-founded Revion, a post-surgery AI visualization tool for breast augmentation. Revion takes in a pre-surgery photo and outputs a personalized, realistic post-surgery visualization.

Women's health has historically been overlooked; it deserves access to the cutting-edge technology that can improve patient outcomes.

Getting the data meant cold-calling clinics and pitching them our solution. Through these calls, I secured data-sharing partnerships with two high-volume breast augmentation clinics.

At Boston College’s Elevator Pitch Competition, I pitched solo to investors and won first place, securing $1,500 from SSC Venture Partners. A couple of months later, our team placed as finalists at BC's Shark Tank.

Education

Boston College

Master of Science in Data Science

Expected August 2027

Chestnut Hill, MA · Cumulative GPA: 4.00/4.00

  • Data Science Program Ambassador
  • Dean's Scholarship
Boston College

Bachelor of Science in Computer Science

May 2026

Minor in Data Science · Chestnut Hill, MA · Cumulative GPA: 3.95/4.00

Contact

Email ↗ LinkedIn ↗ GitHub ↗
Back to top ↑