July 2022 Summaries

The Code Jam World Finals of 2022 marks the culmination of a competitive programming journey where top programmers from a pool of tens of thousands will compete for the title of Code Jam World Champion and a grand prize of $15,000 USD. Scheduled for August 5th, the event will be livestreamed, showcasing the talents of these finalists as they tackle challenging algorithmic problems. Participants, who include returning finalists and newcomers, share personal anecdotes, favorite coding music, and insights into their handles, offering a glimpse into the diverse personalities and experiences behind the competitive spirit. The event promises to be an exciting display of problem-solving skills, creativity, and community engagement, as viewers have the chance to watch live coding demonstrations and hear expert commentary from the Code Jam team.

RandoTek is an initiative launched by members of the Google Developer Groups (GDG) in Beja and Sousse, Tunisia, to address the limited access to technology in remote areas by training young developers. Spearheaded by volunteers like Alaedeen Eloueryemmi and Yasmina Rebai, the program originated when they realized many attendees at a DevFest in Beja struggled to follow a talk due to unfamiliarity with Google technologies. Since February 2022, RandoTek has conducted workshops on technologies such as Flutter, Go, and TensorFlow, reaching over 412 developers across eight regions and five cities in Tunisia. The initiative aims to foster a robust community of tech enthusiasts who can inspire others and potentially form new developer groups in their areas, thus expanding the reach and impact of technology education in the country. Participants express a strong desire to learn more and join GDG groups, highlighting the program's success in motivating individuals to enhance their skills and become future leaders in the tech community.

In the ongoing Serverless Migration Station mini-series, Google Cloud aims to help developers modernize their serverless applications by upgrading language runtimes and transitioning from legacy App Engine APIs to standalone Cloud services. Module 15 specifically focuses on integrating App Engine's Blobstore service into a sample app, which aids in handling user uploads and storing large files, before transitioning to Cloud Storage in the subsequent module. This process starts with a Python 2 app using the webapp framework, which will later be modernized to use Flask and Cloud NDB while moving from Python 2 to Python 3, ultimately making the app more portable across various platforms. The module demonstrates enhancing a sample app by allowing users to add file artifacts during their visits, with new handlers to support this feature, while also preparing users for the transition to Cloud Storage. Although recent updates allow continued use of Blobstore in Python 3 apps, the series encourages migration to Cloud Storage for improved portability, with resources available through codelabs and video guides.

During the second quarter of 2022, Google Machine Learning communities around the world engaged in various educational and innovative activities. TensorFlow User Groups (TFUG) in locations such as Agadir, Taipei, New York, Turkey, and Mumbai hosted events ranging from introductory courses in Python and TensorFlow to advanced discussions on Neural Radiance Fields and Vision Transformers. Notable contributions included real-time SKU detection solutions, overfitting reduction techniques, and Android app integration with TensorFlow models. JAX/Flax specialists explored lightweighting solutions and the implementation of Neural Radiance Fields. Kaggle recognized several Machine Learning Google Developer Experts (ML GDEs) for their contributions with the Google OSS Expert Prize, while publications like "The Kaggle Book" provided strategic insights for competitions. The use of TensorFlow Extended (TFX) was highlighted in various workshops, focusing on deploying machine learning models into production environments. Cloud AI discussions centered around the integration of Vertex AI with MLOps systems, emphasizing the automation of model production. Research initiatives included discussions on Google's advancements like the PaLM model, illustrating ongoing efforts to push the boundaries of machine learning capabilities.

Melissa from Greenville, South Carolina, is highlighted for her app BringFido, which aids pet owners in finding pet-friendly accommodations, eateries, and attractions, showcasing her journey from idea inception to a successful business. The blog post also features international game founders like Arnaud from France, who created the popular app Akinator with his team, Daigo from Japan, known for his award-winning indie games, and Keerti and Kashyap from India, who developed Hitwicket Cricket Games, captivating millions globally. These stories are part of the #WeArePlay series, celebrating the diverse community behind apps and games worldwide, with more stories available at g.co/play/weareplay.

Campaign Anomaly Detector is an unofficial, open-source tool developed by gTech team members at Google, designed to enhance the capabilities of the original Account Anomaly Detector by allowing advertisers to define and detect metric anomalies in their campaigns. Available on GitHub, this solution encourages community contributions to its ongoing development. The tool offers features such as selecting entities to monitor, defining time periods for comparison, choosing key metrics and thresholds for anomaly detection, and setting up email alerts for identified anomalies. It integrates with Google Sheets for configuration and uses a Data Studio dashboard for visualizing results. Although experimental and unsupported, it is intended for illustrative purposes without any warranty, and users must assume all associated risks when deploying it.

Google's annual Solution Challenge invites the Google Developer Student Clubs (GDSC) community to create solutions for real-world issues using Google technologies, focusing on the United Nations' 17 Sustainable Development Goals. This year's contest culminates in a Demo Day on July 28th, where the top 10 finalists will present their projects live on YouTube, and the top three will be selected as grand prize winners. The diverse range of projects includes applications aimed at improving health and well-being, supporting gender equality, enhancing quality education, reducing inequalities, and promoting sustainable communities. Some notable projects are BloodCall from Greece, which facilitates blood donations; Blossom from Canada, which provides menstrual education; Gateway from Vietnam, which offers a COVID-19 digital check-in system; and SIGNify from Canada, which aids communication between deaf and non-deaf individuals through sign language. Each project leverages a variety of Google technologies such as Firebase, Flutter, and TensorFlow, demonstrating innovative approaches to addressing global challenges.