2026 Summer Internship Program: Audio Processing & Machine Learning Intern

Takeda • Boston, MA

3d • $21 - $46 • Hybrid

About The Position

The Quantitative Sciences (SQS) team within Takeda leverages advanced analytics, data science, and AI/ML approaches to enhance drug development. Our team collaborates closely with clinical and translational research groups to design, refine, and validate innovative methodologies that improve outcome assessment and accelerate the evaluation of new therapies. We are currently developing cutting-edge tools to objectively measure symptoms, bridging clinical needs with state-of-the-art computational techniques. As a Machine Learning & Audio Processing Intern, you will contribute to the development of a novel vomit and retch detection algorithm for clinical audio data. This role provides hands-on experience in machine learning, audio signal processing, and real-world clinical applications, while also offering exposure to regulatory and compliance standards for AI in healthcare.

Requirements

• Must be pursuing a Master’s or PhD’s degree in Computer Science, Data Science, Electrical Engineering, or a related field.
• Proficiency in Python and machine learning libraries (e.g., TensorFlow, PyTorch).
• Knowledge of audio signal processing and dataset curation techniques.
• Familiarity with GitHub or similar version control systems.
• Strong problem-solving skills, attention to detail, and interest in healthcare applications of AI.
• Ability to work both independently and as part of a collaborative team.
• Must be authorized to work in the U.S. on a permanent basis without requiring sponsorship
• Must be currently enrolled in a degree program graduating December 202 6 or later
• The internship program is 10- 12 weeks depending on the two start dates ( May 26 th -August 14 th or June 15 th - August 21 st )
• The intern must be able to commit to one of these time frames
• Able to work full time 40 hours a week during internship dates

Responsibilities

• Extract and organize patient audio clips from pre-annotated datasets.
• Integrate synthetic and crowd-sourced audio datasets into the training pipeline.
• Assist in training and refining machine learning models (e.g., EfficientNet, AST) for audio classification.
• Develop and test noise resilience strategies to address interference from environmental sounds.
• Conduct validation studies, including sensitivity, specificity, and cross-validation analysis.
• Maintain detailed documentation of datasets, code, and algorithm performance.

Benefits

• Paid sick time
• Civic Duty paid time off
• Participation at company volunteer events
• Participation at company sponsored special events
• Access to on-site f itness c enter (where a vailable )
• Commuter Benefit: To offset your work-commute expenses, Takeda provides U.S. employees with a fixed monthly subsidy to be used for either public transportation (transit) or parking.

What This Job Offers

• Job Type: Full-time
• Career Level: Intern
• Number of Employees: 5,001-10,000 employees