Transitioning into an AI-first business offers a handful of benefits. According to TechStrong Research, AI can significantly boost productivity, with 60% of developers reporting increased efficiency and 42% noting gains in testing and QA. Slack’s research shows even higher productivity benefits, with 81% of workers experiencing improvements due to AI.
While chatbots are a popular entry point into AI, the real transformative power lies in how you manage and utilize your internal data. But before businesses can capitalize on an AI-first strategy, the foundational steps must be in place. To prepare, organizations will require smart data management methods, as well as a modern approach to integration.
Arguably a great hidden value lies in API connectivity, which could be bolstered with the right partners in place. Paired with the right technology investments in these areas, APIs play a crucial role in enabling this transformation, opening up programmatic access to data, offering a pathway to adaptive AI-based architectures, and unlocking new revenue streams in the process.
Modernize Data Management Practices
As the saying goes, “garbage in garbage out.” An AI-first strategy truly relies on quality data at its core. Yet, data management practices are still unevolved within most enterprises. Vast data lakes are often unmanaged and comprised of various file formats and database styles. Analysts estimate as much as 80-90% of data is unstructured.
It requires terabytes upon terabytes of data to train today’s advanced AI models. And while most large organizations have this data, data management practices are still immature, making it challenging to craft bespoke AI. “The hyperabundance of accessible data has powered today’s surge in AI adoption and generative AI capability,” writes Jozef de Vries, chief product engineering officer at EDB, for CIO.com. “Collecting, cleaning, organizing and securing that data for AI and machine learning have become a project in itself.”
As such, the first action to perform when transitioning into an AI-first company is typically modernizing your data management. Such data preparation is foundational to strategic reinvention using what Accenture calls a digital core, described as “the engine that enables your company to drive reinvention with transformative technology.”
There are many ways to modernize data management, such as inventorizing your estate, rationalizing databases, data sanitization and heightened governance. However, a focus on integration is arguably the most important and often overlooked element. Streamlining the way data is accessed not only internally but also by external developers and machines is paramount to participating in the AI-first economy at large. This is handled via APIs, the lingua de franca of modern digital business.
Recognize the Value of APIs
APIs are the doorway to innovation, unlocking the full potential of data. And their use across industries continues to climb. Impressively, The 2022 API Security Trends Report found the average number of APIs in use to be 15,564 in large enterprises. Salt Security charted a 167% increase in the count of APIs over the past year.
An API-driven strategy can grant both direct and indirect value. For instance, API-first microservices can make internal components more reusable, reducing effort and costs. Data-sharing through APIs also increases collaboration and productivity. You can also reap direct benefits through API-based partnerships or using APIs as a core revenue stream. For example, Harvard Business Review reported that eBay generates 60% of its revenue through APIs.
In the context of AI and large language models (LLMs), APIs have many roles to play, such as opening access to certain AI functions, accessing real-time data and informing model training. As such, the value of AI is intrinsically tied to APIs. “The API ecosystem is undergoing rapid and significant transformation, characterized by two main trends: The proliferation of API providers and the rise of AI-driven APIs,” says Eyal Solomon, CEO of Lunar.dev, on Nordic APIs.
Leverage the Right Integration Tools
With the advent of AI and LLMs, an entirely new range of opportunities are opened up for organizations to get value from their APIs. Yet, taking advantage of this new paradigm will hinge on having the proper tools to maximize an API investment. For instance, the developer portal with which developers integrate APIs should be well-thought out, and designed with quality developer experience in mind.
Typically, an organization’s internal software teams are busy just trying to keep up, and new trends like machine learning and APIs might put them over the edge. As such, organizations often turn to subject matter experts or third-party solution providers who can guide and help drive business value via API development and API integration with LLMs.
For instance, plenty of platforms exist for full lifecycle API management, offering a one-stop shop for everything from API design to development and testing. Alternatively, fit-for-purpose tools exist for areas like documentation and developer portals, SDK generators, API monitoring or API monetization. Partners can also aid in determining where value lies in data, and what could be used in conjunction to boost AI efforts.
Monetize APIs for New Revenue
Particularly interesting in the context of AI is monetization. While not all APIs are priced, the monetization of APIs and data sets through subscriptions or usage fees is already an important avenue of revenue generation for an organization. Interestingly, 65% of respondents to Postman’s 2023 State of the API Report said they make revenue with their APIs. 43% of these respondents said APIs generate more than a quarter of their business’s total revenue.
Where generative AI is concerned, API monetization could be the determining factor in granting an automatic and consistent cash flow to proprietary data. Part of the reason is that AI might soon be the primary consumer of APIs. “We’re not far from applications that self discover and integrate to 3rd party services,” says Sagar Betchu, Co-Founder and CEO at Speakeasy on The API Economy.
To be agile in this future economy of autonomous agents (Gartner refers to this as “autonomic systems“), it will require not only high-quality datasets but a programmatic layer complete with knowledge of its inner workings. Because once an AI determines which API to call given a prompt, it must learn how to construct these requests, communicate them, and parse responses automatically.
Developer portals can play an important role here in a few ways. Firstly, they can aid model training by offering developer resources in the form of reference documentation, walkthroughs, and sample code, improving the accuracy of responses. Secondly, they could provide machine-readable instructions to AI agents on how to automatically integrate, perform read and write operations, and potentially even charge the end user for access.
AI and APIs for the Win
Transitioning to an AI-first strategy with an API-centric approach promises substantial benefits for businesses. To fully realize these advantages, it’s essential to follow a structured roadmap that emphasizes increasing API value and fostering collaboration within your teams. This approach ensures both successful API implementation and a positive return on investment.
However, such a transition demands a significant investment in integration and the expertise of API professionals who can identify and harness the specific ways APIs can deliver value, thereby accelerating the AI transition. Accenture’s guidance highlights the importance of addressing inefficiencies and reallocating resources toward innovation.
Still, navigating the AI landscape presents its own set of challenges. With numerous competing AI services and the propensity of LLMs to produce unreliable outputs, businesses must be cautious. Security risks and code quality concerns further complicate this process. Therefore, having expert partners who understand both the benefits and limitations of AI technology is crucial.
Overall, a strategic investment can enhance API programs and integrate AI/LLMs more effectively. Leveraging API partnerships and expert guidance can help businesses navigate these complexities, helping early adopters maximize the value of their AI-first transitions.