Sharing our story on machine learning at GCC Meetup Philippines

We were invited to share our work in machine learning at the Google Cloud Connect Meetup in the Google Philippines office last Monday, and we were ecstatic. Symph has always been proud of how we explore new technologies to serve our clients better.

A Google Cloud Partner is a company endorsed by Google that can use Google's technology to build better products and services. Symph became a partner in 2010, when we helped businesses use G Suite for their office and email workflows. We were there to talk about one of our projects in the Philippines that used machine learning with TensorFlow.

The project was Geostore, a platform built to monitor roadwork projects funded by the government. It keeps all the information collected during road implementation and audit in one place, so agencies can manage the data better and the public can access it in the future.

The machine learning came in with the information collected for Geostore. The agencies were getting thousands of datasets with project details and, most of all, photos of the roadwork they were auditing, and they had no efficient way to sort through all those photos. One roadwork project could have as many as a thousand photos, so across thousands of projects there were millions of images to sort and classify. Adding machine learning to the system let us sort the road images by the type of road surface in each one, or by whether there was a road in the image at all.

"Google Cloud Connect* is the official online community for G Suite administrators. Google Cloud Connect enables administrators to get critical, real-time updates about G Suite, see what other G Suite customers are experiencing when it matters most, and share best practices."

At the meetup we shared our story with the other Google Cloud Partners and with IT managers of companies in the Philippines that use the Google Cloud Platform. We were humbled by the opportunity.

Albert (left) and Dave (right) telling our story. Photo: https://cdn.prod.website-files.com/67ca96d09e7c183c26e5543f/67cab4644fa0fb094186efd0_634fd398cf993dd75110d10a_Screen%2520Shot%25202022-10-19%2520at%25206.37.25%2520PM.png

There are machine learning opportunities in a wide range of industries. Ours came about because our team did not like clicking through road images to classify them by hand, and they knew machine learning could do it much more efficiently. Some image processing use cases we thought of are checking whether an image is inappropriate, identifying the type of crops or agriculture in satellite imagery, healthcare diagnosis in image-based laboratories, quality control in agriculture, and detecting spoilage from camera images in supermarkets. These are ideas off the top of our heads, and machine learning could be applied to many other areas.

We're glad we got to share how we use this technology, and we thank the GCP community for having us at their Google Cloud Connect meetup. We hope to keep applying these technologies in the work we do.

If you have an idea that could use this technology and you would like to work with us, email us at info@sym.ph and we'd be happy to chat.