Cloud Solutions
We design software to make the most of modern public cloud platforms.
What is Cloud
Cloud computing allows companies to rent infrastructure instead of owning it. Businesses can scale their usage up or down as needed, paying for server use per second. AWS (Amazon), Azure (Microsoft), and GCP (Google) are the leading providers in this space.
What Can the Cloud Do Today
Beyond hosting servers and providing basic infrastructure, modern cloud platforms offer advanced services like databases, analytics, machine learning, and ready-to-use AI models. These platforms now act as universal marketplaces where diverse systems are operated and offered. Many businesses depend on these services to manage data and applications. For some, the cloud is essential to stay competitive.
Which Cloud Platform Should Small Companies Use for Analytics
Looker (formerly Data Studio) from Google is straightforward to set up and uses real pay-per-use pricing that depends solely on the computing resources needed for analysis. It incurs almost no costs when idle, avoiding server fees or per-user monthly charges, making it ideal for startups or exploratory projects. Looker integrates seamlessly with GCP and BigQuery and supports various other data sources.
Power BI from Microsoft is well-suited for larger companies that already have Power BI subscriptions and experience with Microsoft technologies. It integrates effectively with the Microsoft stack and provides numerous features tailored to enterprise needs. It is a natural choice for analytics projects run by larger companies or corporate departments.
Is Data-Driven Development Different from Regular Programming
Data-driven development relies heavily on experimentation, setting it apart from traditional programming. Business expectations, cases, functionality, and data sources are iteratively discovered rather than predetermined. Rapid experimentation cycles are crucial for success, often requiring cloud technologies to achieve results. Because of that data-driven solutions are often developed as cloud solutions.
Should My Developers Be Certified in Cloud? Which Certification?
Having relevant certifications is generally beneficial, even if they have expired. Certifications show that a developer has explored all major features of a cloud platform, not just those used in previous projects. However, certifications can be time-consuming, and a developer with too many certifications may have prioritized earning badges over hands-on work.
Cloud platforms offer diverse tools and technologies. Certifications should match the job requirements. A certified data analyst, for example, will have very different skills than a DevOps engineer or cloud architect.
Should the Developer Know All of the Cloud?
Developers typically have hands-on experience with only a small fraction of what cloud platforms (and computing technology in general) offer. A developer may spend an entire year productively using just a handful of cloud features among hundreds or thousands. The technology landscape evolves quickly, making it nearly impossible to stay current on everything. A good developer remains aware of industry trends, stays updated at a high level, and learns specific topics as they become relevant.
How to Conduct One-Off Data Analysis
One-off data analysis or reporting often involves gathering, filtering, and preparing data, even if the results are used just once. The effort required can be comparable to developing a full analytical product intended for long-term use.
These projects typically demand rapid experimentation, supported by flexible dashboards that can be modified with a few clicks and immediately applied to large volumes of live data for testing results. Such experimentation often leads to spikes in resource usage, where clusters of servers analyze substantial datasets to explore yet another potential pattern. This approach differs significantly from the predictable workflows needed for scheduled analytical reporting, which prioritize deployment stability and controlled resource consumption. As a result, cloud solutions have become the standard for conducting one-off analyses.
Google Looker is a strong candidate for one-off analysis projects, especially when used with BigQuery from GCP. BigQuery offers affordable data storage and the ability to span large server clusters to execute complex queries, while incurring no costs during idle periods.
Exiting the Cloud
We design software solutions to operate beyond public cloud environments.
Public and Private Clouds
Public clouds are universally accessible services, provided by major vendors such as AWS (Amazon), Azure (Microsoft), and GCP (Google). Private clouds, on the other hand, are smaller services either offered by vendors (including the public cloud providers) for dedicated use by a single client, or operated internally by large enterprises. The latter are commonly referred to as 'on-premise' deployments.
Why Are Companies Exiting the Cloud?
Companies are moving away from the cloud for a variety of reasons.
Smaller organizations often cite cost concerns. Setting up a local server can be more economical than renting cloud services. Many advantages of a managed cloud environment may be less relevant to tech-savvy startups that are comfortable with deploying (refurbished) servers.
European businesses often find it easier to achieve and maintain compliance with EU GDPR regulations and industry-specific standards when processing data on company-owned servers within the EU.
Smaller vendors are also influenced by their clients; large corporate clients may dictate preferred cloud providers or push vendors towards exiting the cloud entirely.
Do Companies Use a Mixed Approach?
Yes, many companies adopt a hybrid strategy, utilizing the cloud for development and experimentation, while transitioning production deployments off-cloud. Data-intensive applications often require extensive experimentation during development. Companies leverage the scalability and resource flexibility of the cloud for this stage. However, once applications move to production, resource consumption becomes more predictable, reducing the cloud's advantages.
How Difficult Is It to Exit the Cloud?
The complexity of leaving the cloud depends on the level of reliance. Portable systems with standard integrations can be relocated as-is (referred to as 'lift-and-shift'), whereas systems deeply integrated with specific cloud services may face significant challenges (known as 'vendor lock-in').
Partial cloud exits are also feasible, where certain systems or datasets remain within the cloud.
How Can BISOT Help?
BISOT specializes in evaluating your cloud infrastructure, identifying dependencies on cloud-specific functionalities, and estimating the effort required to transition systems to operate off-cloud.