Y Combinator-backed startup Frekil has been launched as a data engine for medical scans. Referred to as “scale AI for medical scans,” this startup helps procure and annotate medical scans — like X-rays, CTs, Fundus images and more 10 times faster with AI assistance.
Frekil claims that while AI is transforming healthcare, the biggest bottleneck is access to medical images from hospitals and annotations by expert doctors. While healthcare AI and life sciences R&D companies spend months and millions collecting medical images and hiring expert doctors, annotations are still done manually, and are slow and expensive.
READ: Document extraction startup Reducto raises $24.5 million in funding (April 29, 2025)
Frekil aims to solve this problem by working with radiology partners to source large-scale imaging data and pairing it with its global marketplace of certified radiologists. The startup has partnered with radiology chains for large-scale imaging data and has a global marketplace of radiologists for rapid, high-quality annotations.
The founders Nikhil Tiwari and Shivesh Gupta have also built powerful, browser-based annotation tools designed for medical AI which are collaborative, benchmark-driven, and FDA-ready. This helps teams go from raw data to clean, annotated datasets within days.
With Frekil, customers can choose annotators from a global marketplace where everyone is benchmarked on every project.
Frekil also allows customers to connect their existing cloud storage, and fetches imaging data securely for annotation using pre-signed URLs and streams it to browsers of annotators. Frekil users can also set custom annotation guidelines, assign scans, and monitor progress.
READ: HubSpot acquires Frame AI, an intelligent conversation platform (December 6, 2024)
Frekil offers a web-based suite for 2D and 3D segmentation, point/polygon labeling, lesion measurement, and advanced visualization which supports X-ray, CT, MRI, ultrasound, histopathology and more. Its powerful AI tooling allows the speeding up of annotation with zero-shot segmentation, 3D brush & eraser, grow/cut tools, threshold tools, and slice interpolation. It also has measures to ensure quality and consensus.
The founders are engineers from IIT Bombay with experience at Stripe, Sony, and Amazon.