The 5 Pillars of IBM’s App Development and Modernization Strategy
Using some GenAI magic (by GPT4o), we find the following 5 topics in the 103 developer-centric sessions at the IBM TechXchange conference. All of these topics focus on demonstrating IBM’s ability to make application developers more productive by eliminating overhead tasks, while at the same time optimizing performance, compliance, and security. Stepping through these five topics one-by-one shows the key pillars of IBM’s developer strategy.
Topic #1: Cloud-native Development, Microservices, and Containers
The most prevalent topic across developer-focused sessions at IBM TechXchange was building cloud-native microservices apps for IBM’s WebSphere Liberty Java runtime. IBM showed how the resulting microservices can run on open source Kubernetes or on Red Hat’s Kubernetes-based OpenShift developer platform. Developers can use Podman Desktop to easily deploy their locally built containers as pods to Kubernetes clusters to prevent the “but it works on my laptop”-syndrome.
My Take
Convincingly showing Java developers the advantages of developing for the Liberty Java framework is critical for IBM as Liberty is a foundational technology that enhances and integrates IBM’s broader enterprise solutions. Liberty enables customers to develop cohesive software solutions that leverage many different IBM products and technologies.
Despite IBM’s focus on the Java language, Big Blue has responded to Python turning into the de-facto standard development language for AI and data applications. Most of the leading AI frameworks and libraries are only available for Python: PyTorch, Keras, TensorFlow, Transformers, Scikit-learn, etc. This makes Java less attractive for machine learning and AI projects. I have seen IBM add Python libraries to their AI pipelines, showing that Big Blue is working on reconciling Python-driven AI microservices and Java-driven enterprise ones. watsonx in fact comes with Python support (demonstrated in this video).
Microservices architectures allow developers to write microservices in different languages and connect them via REST API, gRPC or event-driven messaging platforms like Kafka.
Topic #2: WebSphere and Liberty Operations, Migration, and Performance
Demonstrating the advantages of migrating to WebSphere Liberty over competing Java frameworks (Oracle WebLogic, Spring Framework by Broadcom, etc.) is critical for IBM’s application modernization strategy. IBM specifically focused on showing Liberty’s performance enhancements, smaller footprint, container support, granular security, and rapid startup times. Combined, these capabilities enable greater scalability in microservices architectures.
My Take
It is key for IBM to show developers that transforming their applications for Liberty is simple and brings significant advantages. IBM’s AI-driven transformation assessment, planning, and automation toolkit (Transformation Advisor) is a critical puzzle piece within this context, as it is crucial to minimize the developer effort of transforming apps for Liberty.
It is important to note that Liberty supports the popular Spring Boot application framework, as this is key to re-assuring the developer community of the open character of Liberty. Allowing developers to take advantage of the Spring Boot ecosystem while having access to Java Enterprise capabilities is important to ensure loyalty.
Topic #3: Java Runtime Technologies
These sessions centered around giving developers the information they need to decide which java runtime technology is suitable for what use case. The focus is on differences in features of IBM supported runtimes such as WepSphere Liberty (IBM’s commercial Java application server), Open Liberty (open source foundation of WebSphere Liberty), Semeru Runtime Open Edition (open source Java runtime without formal certification testing), Semeru Certified Edition (version of Semeru that has passed certification testing and includes an IBM license), and also Spring Boot (supported by Broadcom).
Differences in a Nutshell (based on analysis by GPT4o):
Licensing: WebSphere Liberty and Semeru Certified Edition require IBM licenses and offer enterprise support, whereas Open Liberty, Semeru Open Edition, and Spring Boot are open-source with community or third-party support.
Certification: WebSphere Liberty and Semeru Certified Edition are fully certified, making them ideal for enterprise applications, while Open Liberty and Semeru Open Edition are more focused on flexibility and performance in open-source environments.
Target Use: Spring Boot is primarily focused on rapid microservices development, while Liberty runtimes (WebSphere and Open Liberty) and Semeru are more suited for enterprise cloud-native applications.
Cloud-Native Features: WebSphere and Open Liberty are optimized for cloud-native environments with modularity and containerization, whereas Spring Boot offers robust microservices development with Spring tools.
My Take
IBM’s education around runtimes is a strategic effort to show its value to clients in explaining how to solve real business problems in today’s complex, cloud-native, open source world of enterprise technologies. Allowing clients to dip their toes into open Java frameworks that are still supported by the IBM product portfolio is critical for Big Blue’s credibility as an open source-centric vendor.
Topic #4: Application Modernization and Tooling
Simplifying application modernization is key for developer productivity. IBM’s Transformation Advisor, and watsonx Code Assitant help developers automate the migration and modernization legacy apps for cloud environments. This includes breaking down monolithic apps into microservices to improve agility, scalability, and portability to enable the fast response to customer requirements.
My Take
Automating the transformation of any type of legacy application into modern distributed cloud native apps makes it likely for these applications to stay within the IBM universe. This is a very good reason for IBM to make application transformation as easy as possible.
Topic #5: AI-assisted Development
watsonx Code Assistant leverages GenAI to accelerate code development within a large enterprise context and for specific use cases. For example, watsonx code assistant for Red Hat Ansible Lightspeed generates Ansible code recommendations from natural language prompts, enabling IT teams to automate tasks like network configuration and app deployment. watsonx Code Assistant for Z uses AI to translate legacy COBOL applications to Java on IBM Z mainframes. watsonx Code Assistant accelerates Python development by generating code recommendations based on natural language prompts, helping developers write quality code more efficiently and reducing the need for context switching between requirements and implementation
My Take
It is important to see IBM offering watsonx Code Assistant for development languages other than Java. Python support is the most intriguing new offering, enabling organizations to leverage an on-premises coding agent with a high degree of transparency and privacy.
Final Thoughts
As we can see from the top 5 developer topics at TechXchange 2024, Big Blue is still Java heavy and focused on convincing customers of the advantages of its application frameworks. At the same time, we have seen a lot of Python code at TechXchange and even a watsonx Python SDK. IBM also offers a wide range of Jupyter notebooks showing customers how to leverage the Granite LLM and integrate it into their workflows for AI-driven apps.
IBM’s focus on using AI and specifically LLMs for a wide ranges of real-world use cases was another key takeaway. The company is laser focused on its hybrid multi-cloud story where customers can decide exactly where applications run and data is located. IBM’s Granite model allows for full transparency in terms of the data used for model training and also in terms of the model’s answers.
Finally, it was great to see IBM focusing on its own strengths instead of imitating Amazon, Microsoft or Google. This is key, because IBM has something the other don’t: tentacles into countless verticals all over the globe.
