Reconstruction Algorithm Devleloper

Philips; Bengaluru; On-site; INR 1,500,000 - 2,100,000; Full time

Job summary

Philips in Bengaluru, India, seeks a Reconstruction Algorithm Developer to design, develop, test and maintain MR reconstruction software. You will translate state-of-the-art research into production-grade C++ and Python, and optimize algorithms for speed and image quality.

You will work with AI/ML components (PyTorch, ONNX, TensorRT, CUDA/cuDNN), collaborate across algorithm, software and systems teams, and contribute to project planning and risk management.

Qualifications

• MTech with 5+ years of relevant experience, or PhD with 2+ years of relevant experience.
• Strong C++ and Python production-grade code.
• Experience profiling and optimizing computationally intensive reconstruction algorithms.
• Experience with AI/ML components and tools such as PyTorch, ONNX, TensorRT, CUDA/cuDNN.
• Ability to collaborate across algorithm, AI/ML, software and systems teams.
• Experience in medical devices or healthcare is preferred; healthcare domain experience is a bonus.
• Knowledge of Waterfall and Agile development methodologies.
• Clear and coherent written communication skills; strong interpersonal communication.

Responsibilities

• Design, develop, test, and maintain MR reconstruction software.
• Translate state‑of‑the‑art academic research into production‑grade C++ and Python code.
• Architect and implement advanced MR reconstruction algorithms.
• Analyze reconstruction algorithms to identify performance improvements.
• Optimize AI/ML components integrated into the Recon pipeline (PyTorch, ONNX, TensorRT, CUDA/cuDNN).
• Collaborate with reconstruction algorithm, AI/ML, software, and systems teams to translate intensive algorithms into efficient implementations.
• Investigate and resolve customer complaints.
• Own and deliver assigned modules as part of the team.
• Support project estimation, planning, and risk management.

Skills

C++

Python

PyTorch

ONNX

TensorRT

CUDA/cuDNN

Performance profiling

Image reconstruction

Education

MTech in related field

PhD in related field

Tools

PyTorch

ONNX

TensorRT

CUDA

Job description

Job Title

Reconstruction Algorithm Developer

Job Description

Reconstruction Algorithm Developer

In this role, you are part of

The MR Recon Software team, a key part of the MR Domain SW department in MR R&D within Philips. MR R&D focuses on developing and enhancing MR products and solutions. We have defined aspirational goals and a “North Star” in line with customers’ most pressing needs. We work on improving (a) Speed, to ensure predictable and reduced cost per exam; (b) Comfort, to improve patient and staff satisfaction; and (c) Confidence, to support first-time-right diagnosis.

The MR Recon team works on new AI technologies and GPU implementations and is exploring cloud technology to meet future clinical needs. Team members have opportunities to grow as domain specialists or technical specialists.

Your Role:

• Design, develop, test, and maintain MR reconstruction software.
• Translate state‑of‑the‑art academic research into robust, production‑grade C++ and Python code.
• Architect and implement advanced MR reconstruction algorithms.
• Analyze reconstruction algorithms and implementations to identify performance improvements through algorithmic complexity reduction.
• Optimize AI/ML components integrated into the Recon pipeline, including PyTorch-based models, ONNX, TensorRT, and CUDA/cuDNN.
• Work with reconstruction algorithm developers and AI/ML, software, and systems teams to translate computationally intensive algorithms into efficient, scalable implementations.
• Investigate and resolve customer complaints.
• Own and deliver assigned modules as part of the team.
• Support project estimation, planning, and risk management activities.

You're the Right Fit If:

• MTech or PhD in biomedical engineering, signal processing, computer engineering, or a related field.
• Experience developing reconstruction or image-processing algorithms: 5+ years with an MTech, or 2+ years with a PhD.
• Strong C++ skills and the ability to develop robust, production‑grade code in C++ and Python.
• Experience analyzing and optimizing computationally intensive reconstruction algorithms and software implementations.
• Experience applying profiling and performance‑analysis techniques to identify bottlenecks and deliver measurable improvements.
• Good to have: experience with AI/ML components and tools such as PyTorch, ONNX, TensorRT, and CUDA/cuDNN.
• Ability to improve end‑to‑end performance while maintaining numerical accuracy and image quality.
• Ability to collaborate effectively across algorithm, AI/ML, software, and systems teams.
• Experience in medical devices or healthcare is preferred; prior experience in the healthcare domain is a bonus.
• Knowledge of Waterfall and Agile development methodologies.
• Clear and coherent written communication skills, including the ability to present data visually.
• Strong interpersonal communication skills.

Minimum required Education and Experience

MTech with 5+ years of relevant experience, or PhD with 2+ years of relevant experience.

How we work together:

We believe that we are better together than apart. For our office-based teams, this means working in-person at least 3 days per week.

• undefined
• Onsite roles require full-time presence in the company’s facilities.
• Field roles are most effectively done outside of the company’s main facilities, generally at the customers’ or suppliers’ locations.
• This role is an office-based role.

About Philips:

• Learn more about our business
• Discover our rich and exciting history
• Learn more about our purpose

We are a health technology company. We built our entire company around the belief that every human matters, and we won't stop until everybody everywhere has access to the quality healthcare that we all deserve. Do the work of your life to help the lives of others.

You may still be the right candidate for this or other opportunities at Philips. Learn more about our culture of impact with care here.

• undefined