Search our Blog
Every neural network is biased. True or false?
Creating non-biased algorithms is a complicated matter and a goal that we’re still far from achieving. To do that, the data has to be bias-free, and the engineers creating these algorithms need to ensure they’re not leaking any of their own biases. Needles to say, that AI tends to reflect human societal prejudices.
IoT influences our reality
There are more than 7 billion IoT-connected devices today, and experts estimate that this number will grow to 10 billion by 2020 and 22 billion by 2025. Asset-intensive enterprises like utilities, oil, gas, energy, manufacturing, and construction are progressively deploying IoT solutions to conduct operations with greater productivity and decreased costs. At the same time, retailers, cosmetics producents, and healthcare providers use it to improve their testing, safety standards, and customer experience.
ML and data-driven approach to maximize your profits
Where data-driven approach and business intelligence can increase sales and savings thanks to past and current data, with Machine Learning and predictive models, we’re approaching the future. Businesses incorporate ML into their core processes for a variety of strategic reasons. ML can deliver benefits such as discovering patterns and correlations, improving customer segmentation and targeting, and ultimately increasing a business's revenue, growth, and market position.
The biggest challenge of Data Science
Without automation of provisioning of AI training environments, testing an idea requires even 6 months of work and a huge budget to meet the compliance regulations. For this reason, good ideas often don't manage to even get to the testing phase.
A SOLID look on AI Booster
Some time ago, Google publically announced the success of the AI Booster project. It is a collaboration built on top of Google Cloud Platform and Vertex AI. The project involved Vodafone and Google, along with other partners.
SDLC for Terraform at scale
Solid Potential DevOps Engineers created terraform-based solution currently supports hundreds of monthly deployments by dozens of Platform/DevOps Engineers from different departments.
AI Platforms with Kubeflow
Solid Potential DevOps Engineers delivered a Kubeflow deployment on Google Cloud’s Kubernetes Engine. The solution includes an array of features, including authentication, scaling, and cost management. This infrastructure-as-code solution gave our customer a unified solution to train ML models and was a big stepping stone toward adopting AI in the company.
Self Service infrastructure
The self-service paradigm is where the teams or managers can get an instance of a configured and work-ready environment by filling out a form or work order. We provided this capability to our customers to request AI training and serving environments based on Google Cloud’s Vertex AI product suite.