Sensedia’s No Nonsense AI

In a world in which AI is every day more present and changing, Sensedia has a very practical approach to AI, based on the aspects of AI that are safe and useful:

• You can facilitate your use of AI platforms, but also govern API consumption by AI constructs (like chatbots today and likely agents tomorrow) using Sensedia as an AI gateway
• When implementing new APIs, you can reduce design time and improve API quality with Sensedia's AI Co-pilot, a model trained on best practices to deliver years of API expertise and success with ease
• At API Run time, you can enhance your API operations in production with AI-assisted governance and monitoring tools, including next best action and intelligent troubleshooting

AI Gateway, Outbound and Inbound

According to Gartner, this is what AI Gateways do: “AI gateways, also known as LLM routers or multi-LLM gateways, manage and secure connections to AI providers. Some AI gateways are based on established API gateway products, since REST APIs are the primary mechanism for connecting to AI providers, while others are purpose-built. Software engineering leaders can use AI gateways to apply security to their organization’s AI usages, such as protection of API keys issued by AI providers, multi-LLM routing, cost visibility and data privacy scanning.”

Well, that works both ways: not just for the AI APIs that your application calls (outbound), but also for the APIs you publish, and are called by chatbots or AI Agents (inbound). You might have published those APIs for B2B interactions (like product or ordering information) before AI stormed into the technology scene, but now that LLMs can easily read and interpret an OpenAPI spec, some more protection is in order.

You might want to make those APIs easily accessible by the AI Agents your business partners are working on, but equally you might want to limit the access to unknown chatbots, and other more sophisticated phishing engines. Remember, threat actors are often faster at leveraging AI than security teams, making proactive protection crucial. And that goes well beyond simple API rate limiting, and well into sophisticated API security (which in Sensedia’s case, uses several AI techniques too).

In all the hype surrounding AI, most AI gateways just focus on outward calls – but that’s just half of the picture. Sensedia’s AI gateway focuses both on outbound and inbound.

Gen AI When You Design APIs

There are a plethora of code assistants out there, aiming at developing new code faster and with better quality: most companies are experimenting with them, with developers tending to use the one they prefer to develop the things they need to.

The market of code assistants is a wild west, changing by the day. In addition to that, there are very few API Design-specific code assistants, most of them requiring extensive training. Sensedia offers an API-first design assistant: it is named AI Copilot, but unlike many others, it guides the developer in a context-based API generation. It is secure, enterprise-ready, and includes API regression test generation. The knowledge and the best practices coming from years of API projects by Sensedia professionals and consulting services are embedded in it. It is different!

 … and there is a Copilot for generating complex integration flows coming in the next months!

AI For APIs in Production

Most of the API management offerings focus on outward API Gateway functionality, some on API design, as per the previous two sections. At Sensedia, we don’t think that’s enough, because AI can be used for running your existing APIs better.

Well run API strategies are still a myth for a lot of companies out there. Everybody recognizes the need to govern APIs, very few are successful in applying just enough governance to design and run their APIs in an agile way, serving the needs of their business. And how many years have we tried to leverage API reuse?

Sensedia is launching an API redundancy detector that uses AI to analyze API contracts, and put duplicates in the spotlight. Developers understandably go down the path of quickest time to results, so as Gen AI code assistants make creating new code easy (see “Gen AI When You Design APIs” above), that might give API reuse even less of a chance. Before you set off developing an API, maybe there is something similar lying in your developers’ portal already?

Improving API contracts is another area where AI can offer practical help. Sensedia has added Smart Recommendations powered by AI to assist and recommend next best actions on that.

In addition, Sensedia is evolving AI-assisted troubleshooting (that requests error summaries and issues recommendations on how to fix them).