Test Automation as a Zero Downtime Deployment Supporting Technique

Posted by Daitan Innovation Team
On March 25, 2022

Zero Downtime Deployment is a strategy to deploy software without interrupting the service, often using methods like canary, blue-green or rolling deployments. These may involve a lot of domains and techniques, but of the most important ones there is testing. In this blog we'll look at the role of manual testing and test automation.

Monitoring and Observability

Posted by Daitan Innovation Team
On March 17, 2022

The way we build systems is changing, and with it, traditional methods begin to fail to handle modern architectures. In this article we are going to discuss monitoring and observability, two concepts that will make the process of solving the most annoying bugs a breeze (or at least a little easier).

Architectural Design Patterns: Cluster Immune System

Posted by Daitan Innovation Team
On February 26, 2022

In architectural design patterns, a Cluster Immune System is an extension of a technique called Canary deployment. Here, solutions used to monitor the production environment are linked to the release process.

UX Writing and Chatbots – Basic Concepts to Start Applying Right Now

Posted by Daitan Innovation Team
On February 18, 2022

Good communication is the foundation for developing successful relationships between companies and their consumers. And to meet the great demand for communication, companies provide a lot of informative content and use several channels or software that automatically responds. Thus, a challenge arises, how to make the communication experience better and more humanized? Today we are going to share some important concepts for those who are starting or need to create some content or chatbot dialogs.

Zero Downtime Deployment Techniques – Canary Deployments

Posted by Daitan Technology Council
On February 10, 2022

A Canary deployment is a zero-downtime deployment technique that allows a gradual transition of traffic from the current version of an application to a new version. The gradual transition of traffic is usually orchestrated using a data-driven approach, with metrics dictating whether to continue or roll back. It is considered an improvement over Blue-Green deployments which implement a complete cut-over of traffic to the new version.

Keeping Data Under (Your) Control with Apache Airflow

Posted by Daitan Innovation Team
On January 28, 2022

Clients are demanding more and more insights based on their data, but to achieve that, some foundational resources need to be in place to allow more advanced analysis and techniques. There are lots of tools and disciplines related to data handling in all aspects, both on premise and as cloud services, but we see that our clients value keeping a certain level of ownership over the entire process. In this context, Apache Airflow pops up as a good candidate for addressing this business need.

Architectural Design Patterns: Infrastructure as Code

Posted by Daitan Innovation Team
On January 28, 2022

Imagine you have dozens of completely different services that require different environments, or maybe you have weekly (or even daily!) deployments. Now think about the amount of work to define and manually manage the configurations for these environments. We can clearly see that this type of activity does not scale well. In this post, we will explore a technique called Infrastructure as Code (IaC), which will help us overcome these challenges.

Principles of Accessible Digital Design

Posted by Daitan Innovation Team
On January 27, 2022

The definition of accessibility is "the quality of being easily understood, used and appreciated". Therefore, when we design something to be accessible, from the construction of a building to a digital product, our goal is that all people, independently of any disability, could easily use the product without experiencing loss.

Efficient Distributed Training in Deep Learning

Posted by Daitan Innovation Team
On January 21, 2022

At the heart of any deep learning system, there is a significant effort in building it. In the last few years, scaling deep learning models has become one of the most promising paths for artificial intelligence, specifically for large AI laboratories like OpenAI and DeepMind. To have a clear feeling of this trend, let us examine recent large language models (LLMs) such as BERT and GPT-3.

Zero Downtime Deployment Techniques – Blue-Green Deployments

Posted by Daitan Technology Council
On December 31, 2021

A Blue-Green deployment is a relatively simple way to achieve zero downtime deployments by creating a new, separate environment for the new version being deployed and switching traffic into it. A rollback happens just as easily, with a traffic switch to the old version. As with any deployment technique, there are advantages and disadvantages to it.

Zero Downtime Deployment Techniques: Rolling Update

Posted by Daitan Technology Council
On December 3, 2021

Deploying a new version of a product to production is one of the most critical moments in the Software Development Lifecycle. It can go from sheer excitement to release the latest features, to a nightmare of cascading failures and outages. In this series of posts, we will explore deployment techniques that can be used to deploy a new version of an application without causing disruption to end users.

Machine Learning for 5G Technology, a Case Study

Posted by Daitan Innovation Team
On November 13, 2021

Identifying signal modulation types using deep convolutional neural networks. In this article, we will look at how the NG-RAN can use an end-to-end deep learning system to recognize and classify modulated RF signals and adjust channel parameters appropriately, so both ends of the communication achieve optimal, effective resource usage.

An Introduction to Federated Learning

Posted by Daitan Innovation Team
On October 26, 2021

How can we train machine learning models using distributed sensitive data? What if we could take advantage of the same properties of crowdsourcing to solve important machine learning problems? That is the idea behind Federated Learning. Federated Learning is a machine learning framework that allows data scientists to train statistical models using sensitive data from users, without uploading user data to servers. It is distributed training technique where training and testing occur locally (at the edge) and only the meta-information is sent to a central server that combines multiple model updates (from multiple clients) into a new model.

GitHub Copilot - A Code Autocomplete Tool on Steroids

Posted by Daitan Innovation Team
On September 8, 2021

Recently, Github and OpenAI released one of the most anticipated AI-based tools for developers — Github Copilot. The Artificial Intelligence (AI) tool is advertised as a pair programming assistant that does much more than usual code autocomplete tools out there. By no means is Copilot a tool intended to substitute developers in any way. Instead, the tool is meant to be used as an assistant that can facilitate many of the “boring” and “repetitive” parts of programming and lets coders worry about parts of the process that require human thinking and reasoning.

A Short Discussion on Bias in Machine Learning

Posted by Daitan Innovation Team
On June 8, 2021

Machine learning systems are being used to decide if someone should receive loans from a bank, send people to jail, or determine the optimal locations where extra police force should be sent. Though many people see these late advances as a good thing (and in fact, they could be), there is a problem that can easily outweigh the possible benefits of each of these solutions. A problem that relates to our deepest social issues and can show up on pretty much any stage of a machine learning project lifecycle.That is the problem of bias.

The Time Series Forecasting Playground

Posted by Daitan Innovation Team
On June 2, 2021

Here at Daitan, time series forecasting has been one of the most important applications of machine learning and today, we are pleased to announce the first (to our knowledge) time series forecasting playground. Inspired by the Neural Network Playground and the GAN Lab, the The Time Series Playground is an interactive open-source tool designed to provide intuition on how to train AutoRegressive Feed Forward Neural Networks for time series forecasting.

The Untapped Potential of AI in Financial Services

Posted by Daitan Innovation Team
On May 27, 2021

To keep up with the evolving finance market, you first need to become savvy in the digital market. This means taking advantage of AI in financial services. It is now emperative to take advantage of your data and leverage the wealth of information within. Emerj has highlighted just how important this is within their in-depth […]

Introducing MLOps

Posted by Daitan Innovation Team
On May 14, 2021

An introductory discussion about MLOps and a set of essential practices for scaling Machine Learning-powered applications.

How to Build a Custom Disaster Recovery Process for AWS Applications

Posted by Daitan Innovation Team
On April 6, 2021

Have you ever considered a scenario in which you lose half of the information that you stored in a cloud folder for your company’s new project because someone mistyped a command?¹ And what about losing the last 2 hours of your team’s chat history because of a lightning strike 100 miles away from you?² Users rarely think about outages in the cloud scenario. Yet, like our home computers, servers can suffer from problems like power grid failure, software issues, or human errors.

How to Build Your Own Intelligent Assistant Using RASA

Posted by Daitan Innovation Team
On November 25, 2020

Chatbots are everywhere. So are the t

Exploring the Viability of Generative Adversarial Networks for Audio Denoising

Posted by Daitan Innovation Team
On November 18, 2020

There are various definitions of audio denoising. For the purposes of this project we interpret audio denoising to be the removal of any sound other than the primary speaker's voice. Thalles Santos Silva covers the mathematical concepts behind denoising and the CNN in his 2019 article. He also provides background information about the two datasets involved (Mozilla Common Voice English dataset and the UrbanSound8k dataset).

Building a Voice Recognition System with PyTorch by Taking Advantage of Computer Vision Techniques

Posted by Daitan Innovation Team
On June 18, 2020

In this piece we describe how we built a reasonably performing Voice Recognition System with PyTorch, using deep learning Computer Vision techniques. With results as good as 90.2% accuracy using different training and testing samples, with only 25% of the original dataset size, we demonstrate how it is currently possible for different AI domains to leverage knowledge from each other to improve their techniques and outcomes.

Privacy-Preserving Data Sharing for Data Science

Posted by Daitan Innovation Team
On April 15, 2020

In the last 2 decades, with the increasing availability of sensors and the popularity of the internet, data has never been so ubiquitous. Yet, having access to personal data to perform statistical analysis is hard. In fact, that is one of the main reasons we, as data analysts, spend so much time doing research using “toy” datasets, instead of using real-world data.

Assessing Audio Quality with Deep Learning

Posted by Daitan Innovation Team
On February 12, 2020

How to train a Deep Learning system to estimate Mean Opinion Score (MOS) using TensorFow 2.0. -- If you’ve ever used VoIP (Voice Over IP) applications like Skype or Hangouts, you know that audio degradation can be a problem. In video or audio conferences, perhaps with clients and prospects, audio quality is important.

The Fundamental Tool That Data Scientists Can’t Miss

Posted by Daitan Innovation Team
On December 20, 2019

How business requirements can prevent you from using available Machine Learning tools and what to do about it. -- When hearing the term Convex Optimization, most people will immediately start talking about how gradient descent is the most awesome thing there is, how we can add momentum to it, how can we choose, adapt, or even circumvent the choice of the step-size parameter, and so on. However, in reality, convex optimization goes well beyond gradient descent and its variants.

How To Build a Deep Audio De-Noiser Using TensorFlow 2.0

Posted by Daitan Innovation Team
On December 1, 2019

In this article, we tackle the problem of speech de-noising using Convolutional Neural Networks (CNNs). Given a noisy input signal, we aim to build a statistical model that can extract the clean signal (the source) and return it to the user. Here, we focus on source separation of regular speech signals from ten different types of noise often found in an urban street environment.

Storing and Retrieving Machine Learning Models at Scale With Distributed Object Storage

Posted by Daitan Innovation Team
On September 6, 2019

The need to quickly create, store, and fetch machine learning models at scale is rapidly increasing. Examples of applications include recommender systems that are based on individual customers’ purchasing habits or detection of fraud attempts based on each customer’s past behavior, among many others.

Leveraging Deep Learning on the Browser for Face Recognition

Posted by Daitan Innovation Team
On August 20, 2019

Face recognition is probably one of the long-awaited technologies of recent decades. From Hollywood movies and TV sci-fi series to actual cell phone solutions, the face seems to be the perfect authenticator. But, despite the hype, the tech didn’t look ready for a long time. However, recent advances in machine learning seem to be worth the wait. To get an idea, let’s take a look at what the big four tech companies are doing in this area.

Exponential Smoothing Methods for Time Series Forecasting

Posted by Daitan Innovation Team
On August 7, 2019

In our last two articles, we covered basic concepts of time series data and decomposition analysis. Following that, it’s now time to apply that knowledge to a practical algorithm. In this piece, we provide an overview of Exponential Smoothing Methods.

A Visual Guide to Time Series Decomposition Analysis

Posted by Daitan Innovation Team
On August 2, 2019

Last time, we talked about the main patterns found in time series data. We saw that, trend, season, and cycle are the most common variations in data recorded through time. However, each of these patterns might affect the time series in different ways. In fact, when choosing a forecasting model, after identifying patterns like trend and season, we need to understand how each one behaves in the series. With this goal in mind, let’s explore two different pre-processing techniques — additive and multiplicative decomposition.

Fundamentals of Time Series Data and Forecasting

Posted by Daitan Innovation Team
On July 25, 2019

Time series forecasting is the use of statistical methods to predict future behavior based on historical data. This is similar to other statistical learning approaches, such as supervised or unsupervised learning. However, time series forecasting has many nuances that make it different from regular machine learning. From data processing all the way to model validation, time series forecasting is a different beast.Many companies are exploring time series forecasting as a way of making better business decisions.

Innovation that Works

Enabling business growth through emerging technology.