Full-Time
Machine Learning Engineer
HOPPR
51-200 employees
Medical imaging data platform for AI
Compensation Overview
$190k - $210k/yr
No H1B Sponsorship
Remote in USA
Remote
U.S. citizens or permanent residents only.
US Top Secret Clearance Required
 AI & Machine Learning (2) |
|---|
| , |
- Bachelor’s degree in Computer Science, Engineering, or a related field with 5+ years' experience in relevant roles
- Proficiency in Python and machine learning frameworks such as PyTorch or TensorFlow
- Strong understanding of MLOps practices, including model deployment, CI/CD pipelines, and performance monitoring
- Experience working with cloud platforms (e.g., AWS, GCP, Azure) and tools like Docker and Terraform
- Exceptional problem-solving skills, ownership mindset, and a collaborative approach
- Thrive in a dynamic and rewarding environment that emphasizes excellence, autonomy, and impact
- Develop, deploy, and maintain state-of-the-art machine learning models for medical imaging, NLP, and multimodal tasks
- Design and implement robust, scalable ML pipelines and shared infrastructure to support agile experimentation and deployment
- Collaborate with researchers to translate novel algorithms into production-ready solutions
- Build and maintain MLOps tools and practices, including automated testing, continuous integration, and monitoring of deployed models
- Optimize model performance for speed, reliability, and scalability in production environments
- Partner with clinicians, engineers, and product teams to align machine learning efforts with clinical and product needs
- Familiarity with healthcare data and clinical workflows is a plus
Hoppr.ai provides a generative AI platform focused on medical imaging. It gives developers access to diverse medical imaging data, pre-configured environments, and the latest AI/ML tools so they can build, train, and test medical imaging models. The platform works by offering ready-to-use datasets and infrastructure in a managed environment, helping researchers experiment and scale model training while maintaining privacy and regulatory compliance. What sets Hoppr.ai apart from competitors is its emphasis on medical imaging datasets combined with privacy safeguards and compliance-ready workflows, enabling safer and scalable AI development in healthcare. The company’s goal is to make it easier and more cost-effective to create healthcare AI solutions that can be used in real-world clinical settings.
Company Size
51-200
Company Stage
Series A
Total Funding
$31.5M
Headquarters
Chicago, Illinois
Founded
2019
Simplify's Take
What believers are saying
- $31.5M Series A from Kivu and Greycroft scales platform through 2026.
- NVIDIA GTC 2026 integration of NV-Reason generates synthetic DICOM datasets.
- TestDynamics Satori partnership accelerates fine-tuned models into health systems.
What critics are saying
- NVIDIA NV-Reason open models commoditize HOPPR foundation models within 12 months.
- Microsoft Azure AI Foundry captures enterprises with integrated cloud services by 2027.
- FDA mandates QMS for all AI platforms, erasing HOPPR compliance edge in 18 months.
What makes HOPPR unique
- HOPPR AI Foundry integrates QMS aligned with ISO 13485 and IEC 62304 for medical imaging AI.
- Curated datasets from 19 million studies across 18 sites ensure data provenance traceability.
- Proprietary foundation models like EB 2D Mammography achieve ROC-AUC 0.9 for cancer detection.
Help us improve and share your feedback! Did you find this helpful?
Benefits
Health Insurance
Dental Insurance
Vision Insurance
401(k) Retirement Plan
401(k) Company Match
Paid Vacation
Parental Leave
Remote Work Options
Hybrid Work Options
Company Equity
Growth & Insights and Company News
6 month growth
↑ 0%1 year growth
↑ 8%2 year growth
↑ 0%Platform control becomes owned infrastructure; screening AI shows real labor proof - march 20, 2026. GE's Intelerad close turns enterprise imaging strategy into balance-sheet reality, while new screening evidence and cyber/helium disruptions sharpen what buyers will actually fund One big thing. Enterprise imaging crossed from conference narrative into owned infrastructure this week: GE's Intelerad close made platform control tangible, new breast-screening AI data made labor relief more credible, and cyber/helium shocks kept resilience at the center of buying decisions. Listen to this week's Marketstrat Pulse Insight: Key takeaways. * Enterprise imaging control moved from messaging to ownership. GE HealthCare's Intelerad close is the week's most material signal because it puts a cloud-first enterprise imaging layer inside an OEM stack, raising bundling leverage, recurring-software exposure, and switching-cost risk for providers and independent AI vendors. * Breast AI finally delivered a prospective labor signal that matters. The Nature Medicine trial showed materially lower radiologist workload with higher cancer detection, but higher recall means the economics are still pathway-dependent. This is evidence for workflow redesign, not effortless margin expansion. * Pediatric imaging remains underbuilt as a commercial and regulatory category. The JAMA labeling analysis and OXOS's pediatric clearance expansion point in the same direction: low-friction workflow tools can move faster than pediatric-specific AI categories that require heavier evidence, labeling, and liability work. * The model layer is commoditizing faster than the workflow layer. HOPPR's NVIDIA integration, Circle's vascular CT expansion, and GE's Springbok collaboration all reinforce the same commercial truth: distribution and installed workflow matter more than open-model novelty by itself. * Reimbursement was quiet, not solved. In the accessible strict-window source set, no material new imaging-specific CMS/MAC LCD/NCD/CPT action surfaced. That keeps enterprise contracting, throughput math, and operational ROI as the primary monetization path for imaging AI. * MIS and medtech resilience still shape imaging budgets. Imperative Care's financing and Stryker's cyber disruption both reinforce that procedural ecosystems and operating resilience remain live competitors for the same hospital capital envelope. Innovation hook. Breast screening AI finally produced prospective labor proof - but not frictionless economics The week's highest-signal evidence item was not another accuracy claim. It was a prospective workflow result. In the Nature Medicine study, the AI strategy cut radiologist reading volume sharply while increasing screen-detected cancers but recall also moved higher. That makes the commercial lesson unusually clear: screening AI can create real labor relief, yet the economic case still depends on what happens downstream. A buyer cannot underwrite the value of fewer reads in isolation if workup volume, callbacks, or pathway complexity rise at the same time. In practical terms, this is evidence for operating-model redesign, not plug-and-play margin expansion. Data basis: Study-reported radiologist readings, cancer detection, and recall indexed against standard workflow. Prospective breast-AI evidence now points to a real labor-capacity lever. The catch is that recall still matters commercially, so the value case is workflow redesign, not just "better detection." Market lens - North America ultrasound market. Care model divergence: Same region, different delivery logic North America should be read as a mature installed-base, workflow- and labor-driven ultrasound region, not as an access-creation market. The region is still overwhelmingly U.S.-driven in scale. By 2035E, the U.S. represents about 91% of North America systems revenue, 91% of ecosystem revenue, and roughly 91% of unit shipments. Canada is much smaller in scale, but it matters strategically because it changes the regional care-model mix and highlights that North America is not one single commercial model. The key takeaway is that the two countries are directionally aligned on technology, but not identical in delivery logic. In both markets, compact / portable becomes the largest revenue engine by 2035E and handheld / POCUS becomes the fastest shipment-growth engine. Where they diverge is in site of care, procurement logic, and monetization intensity. Signal Pulse heatmap - mar 14 - 20, 2026. Event-level The heatmap shows a week where platform control and clinical evidence shared the top tier. GE's Intelerad close and the new breast-AI trial both scored at the structural end of the range because they affect how buyers think about ownership and labor. Mid-tier signals clustered around financing, governance, quantification, and operational reality checks. The pattern matters: the market did not reward raw algorithm novelty this week. It rewarded events that either changed control of the deployment surface or improved clarity on whether AI actually alters operating models. That is a more mature signal pattern than the one-off clearance-heavy weeks seen earlier in the year. The strongest signals this week were ownership and operating-model signals, not feature releases. Platform control and evidence quality dominated the score distribution. Quick-glance table. This week's Pulse research note also covers: GE HealthCare's Intelerad close and the shift from enterprise imaging strategy to owned infrastructure; breast screening AI evidence and the real economics of labor relief versus downstream recall burden; FDA and governance signals including pediatric imaging gaps, portable X-ray expansion, and MR-guided breast biopsy workflow advances; platform-control moves across imaging IT, open-model tooling, and vascular CT analysis; MRI operational risk and medtech resilience, including helium sensitivity and cyber-related disruption; minimally invasive surgery funding and procedural ecosystem competition; and provider, OEM, payer, and AI-vendor read-through across workflow, distribution, and capital allocation. About Marketstrat Marketstrat(R) is a market intelligence and GTM enablement firm committed to empowering clients in data-driven industries. Under the Markintel(TM) brand, it delivers robust market intelligence, while GrowthEngine solutions offer specialized GTM advisory and app-based tools - together fueling growth, innovation, and competitive advantage. For more information, visit www.marketstrat.com. Marketstrat(R) is a registered trademark and Markintel(TM) is a pending trademark of Marketstrat.
HOPPR has integrated NVIDIA's NV-Reason and NV-Generate open models into its AI Foundry platform for medical imaging development, announced at NVIDIA GTC 2026. The HIPAA-compliant platform combines accelerated computing, curated datasets and foundation models for building imaging AI applications. NV-Reason provides multimodal reasoning for chest X-ray interpretation, generating structured analytical steps alongside outputs for greater transparency. NV-Generate creates synthetic DICOM imaging datasets to support model training and validation where real-world data is limited. Built on NVIDIA A100 and H100 GPUs, the Foundry enables developers to train and fine-tune medical imaging models using optimised infrastructure. HOPPR's Forward Deployed Services offers expert support for model customisation, combining machine learning engineers, data scientists and clinical experts to refine imaging applications.
HOPPR to debut purpose-built AI development platform for medical imaging at RSNA 2025. HOPPR, a Chicago-based company focused on medical imaging AI development, will launch the HOPPR(TM) AI Foundry at the Radiological Society of North America (RSNA) 2025 Annual Meeting. The platform addresses a persistent challenge in medical imaging AI: developers have historically faced a choice between innovation speed and regulatory compliance. The Foundry combines proprietary foundation models, curated datasets with established provenance, and an integrated Quality Management System designed to support both rapid development and regulatory preparation. The platform will debut as healthcare organizations increasingly seek validated AI tools beyond proof-of-concept pilots. HOPPR positions the Foundry as the first purpose-built AI development platform for health imaging accessible to developers that operates under a Quality Management System aligned with ISO 13485, IEC 62304, ISO/IEC 42001, and ISO 14971 standards. Foundation models meet healthcare's compliance framework. The HOPPR(TM) AI Foundry integrates vision transformer-based foundation models built using self-supervised learning, enabling developers to adapt models across classification and anomaly detection tasks through intuitive workflows. The platform offers access to one of the largest, curated medical imaging datasets in the private industry, comprising over 19 million studies from 18 partner imaging sites, according to the company. "The HOPPR(TM) AI Foundry represents a breakthrough in innovation infrastructure for trustworthy, scalable AI in medical imaging. Developers no longer need to choose between innovation or compliance." - Khan Siddiqui, MD, CEO and Co-Founder, HOPPR "The HOPPR(TM) AI Foundry represents a breakthrough in innovation infrastructure for trustworthy, scalable AI in medical imaging," said Khan Siddiqui, MD, CEO and Co-Founder of HOPPR. "Developers no longer need to choose between innovation or compliance. The HOPPR AI Foundry is built to accelerate progress, from experimentation to real-world readiness, while maintaining the traceability, documentation, and quality controls required to enable regulatory compliance." The platform's integrated QMS provides version control, traceability, and documentation to support lifecycle management - capabilities that typically require separate infrastructure investments. Developer-friendly APIs allow teams to train, evaluate, and embed AI models directly into imaging workflows or partner applications. Data provenance as development infrastructure. Data quality and provenance remain persistent challenges in the development of medical imaging AI. The Foundry addresses this through curated datasets with known origins, allowing developers to either use HOPPR's labeled and validated datasets or bring their own data for model fine-tuning. This flexibility targets different development scenarios, from early-stage experimentation to production-ready validation. The platform's architecture separates it from AI orchestration platforms like deepcOS and marketplace ecosystems such as CARPL.ai, which focus on deploying and monitoring existing AI models in clinical workflows. HOPPR instead provides the upstream development environment where models are built and refined before deployment - filling a distinct infrastructure gap in the medical imaging AI ecosystem. Strategic context: the AI development infrastructure market. The medical imaging AI market has evolved from narrow point solutions addressing specific clinical tasks to more comprehensive platforms. While companies like Aidoc and DeepHealth focus on clinical AI applications and deployment infrastructure, and technology giants including NVIDIA and Google Cloud offer broad AI development tools applicable to healthcare, HOPPR's Foundry occupies a more specialized position: a development platform purpose-built exclusively for medical imaging that embeds regulatory compliance from the start. Project MONAI, the open-source framework founded by NVIDIA and King's College London, provides domain-optimized tools for medical imaging AI development. HOPPR differentiates by offering a managed platform with integrated compliance infrastructure and proprietary foundation models alongside the tooling, targeting organizations that require both development acceleration and regulatory preparation support. "For us, the HOPPR AI Foundry is transformative," said Sham Sokka, Chief Operating and Technology Officer at DeepHealth. "We can leverage our data and clinical expertise to fine-tune models, rapidly test in production, and then scale as we see improved outcomes. Foundry is instrumental in the holy grail of adaptive learning in clinical workflow." Clinical validation and market access challenges. The Foundry's emphasis on integrated quality management systems addresses a documented gap in medical imaging AI adoption. According to research presented at industry conferences, the medical imaging AI market remains fragmented with numerous unproven solutions, and many certified radiology AI tools are approved but not independently validated for clinical outcomes, workflow integration, and governance requirements. The platform's approach to adaptive learning - enabling models to evolve based on real-world performance while maintaining traceability - targets what DeepHealth's Sokka described as the "holy grail" of clinical AI deployment. This capability becomes increasingly relevant as healthcare organizations move beyond pilot programs to production-scale implementations requiring ongoing model refinement. Developer access and implementation. HOPPR is exhibiting the Foundry at RSNA in South Hall Booth #4000, with demonstrations of the platform's fine-tuning workflows and evaluation tools. The company's mission centers on democratizing access to responsible AI development in medical imaging by removing infrastructure, data management, and compliance barriers that typically require significant capital and expertise investments. The platform's architecture reflects lessons from early medical imaging AI deployments, where developers often discovered regulatory and quality management requirements late in development cycles. By embedding these frameworks from the beginning, HOPPR aims to compress the timeline from concept to regulatory submission - a critical factor for smaller organizations and academic developers without dedicated regulatory infrastructure. The medical imaging AI landscape continues to evolve rapidly, with platforms like Microsoft Azure AI Foundry introducing multimodal healthcare models and orchestration capabilities. HOPPR's focused approach on medical imaging development infrastructure with integrated compliance represents a bet that specialized tools purpose-built for healthcare's regulatory environment will enable faster, safer AI adoption than general-purpose platforms adapted for medical use. For healthcare organizations evaluating AI development strategies, the Foundry offers an alternative to building internal infrastructure or adapting general-purpose tools - provided the proprietary platform approach aligns with their development workflows and strategic priorities.
As part of its commitment to make state-of-the-art AI solutions available to health systems, TestDynamics has partnered with HOPPR to offer a secure development infrastructure for medical imaging within a quality management system, traceability of data provenance, and version control as part of its Satori AI Platform.
hoppr.ai | Materna Medical, a Mountain View, Calif.-based women's pelvic health company, raised $20m in Series B2 funding from GLIN Impact Capital, Wealthing VC Club, and Citrine Angels.