{"id":125697,"title":"Hundred: Automated inspection for precision hardware parts","tagline":"Catch sub-100μm visual defects before parts reach the assembly line","body":"Hey everyone, we’re[ Erica](https://www.linkedin.com/in/erica-diaz1/),[ Iris](https://www.linkedin.com/in/iris-liao-03b041152/), and[ Imani](https://www.linkedin.com/in/imani-gosserand/), co-founders of[ Hundred](https://hundred.systems/) \n\nHundred replaces the typical human visual inspection done on precision hardware parts with an automated system that can scan a whole shipping tray of parts at once.\n\n<https://youtu.be/xC9_YnK2PJc>\n\n**The Problem**\n\nSmall part designs change too fast for traditional machine vision to keep up, so factories still inspect parts by hand: literally people with microscopes sorting out bad parts. As manufacturers move away from China to the US or India, operators familiar with common defects don’t move with them. While trained operators miss \\~15% of defects, untrained operators miss even more. \n\nOne bad part can turn a finished product into scrap material and more than 3 bad parts in an hour can stop the assembly line. Defects cost U.S. manufacturers $45B a year, and line downtime costs another $245B.\n\n**What we’re building**\n\nHundred eliminates human subjectivity by imaging parts at mass production volumes, right in their shipping trays. A full tray can be placed in the Hundred rig, which detects defects smaller than 100µm on a range of materials and finishes. Where human inspectors might require a month of training, Hundred can stop defects on Day 1.\n\n**Our goal is to make sure engineers can set their specs once and know that part quality will hold up when they leave the factory.**\n\n**The Team**\n\n**Erica** designed and ramped consumer hardware at Meta, Google, and Magic Leap for 5 years where she root-caused device failures to bad parts and drove vendors to fix them.\n\n**Iris** was a Quantitative Researcher at Citadel. She then spent a year at Imperial College London researching ML on 3D meshes and other structured data.\n\n**Imani** worked for 5 years at Twitch as a software engineer at the intersection of notifications, short-form content, and ML for hundreds of millions of streamers and viewers.\n\n**Our Ask**\n\nWe’d love to connect with precision hardware companies and contract manufacturers who are struggling with part inspection or looking to scale their inspection capacity.\n\n**Learn more at [hundred.systems](http://hundred.systems)**","slug":"WhN-hundred-automated-inspection-for-precision-hardware-parts","created_at":"2026-11-03T15:00:00.128Z","updated_at":"2026-11-18T06:22:46.379Z","total_vote_count":8,"url":"https://www.ycombinator.com/launches/WhN-hundred-automated-inspection-for-precision-hardware-parts","share_image_url":"//bookface-static.ycombinator.com/assets/ycdc/yc-og-image-c440a0ad1dacfb86eeeb343717479cc54d256614449b4ef719977a0a451f8bc8.png","company":{"id":37732,"name":"Hundred","slug":"hundred","url":"http://hundred.systems/","logo":"https://bookface-images.s3.amazonaws.com/small_logos/cb93bf36f8b635434211fe78f5b4192e14187067.png","batch":"Fall 2026","industry":"Industrials","tags":["Hard Tech","Hardware","Machine Learning","Computer Vision","Industrial"],"search_path":"https://bookface.ycombinator.com/company/37732"}}