Home News Daniel Ciolek, Head of Research and Development at InvGate – Interview Series

Daniel Ciolek, Head of Research and Development at InvGate – Interview Series

0
Daniel Ciolek, Head of Research and Development at InvGate – Interview Series

Daniel is a passionate IT skilled with greater than 15 years of experience within the industry. He has a PhD. in Computer Science and an extended profession in technology research. His interests fall in multiple areas, reminiscent of Artificial Intelligence, Software Engineering, and High Performance Computing.

Daniel is the Head of Research and Development at InvGate, where he leads the R&D initiatives. He works together with the Product and Business Development teams to design, implement, and monitor the corporate’s R&D strategy. When he isn’t researching, he’s teaching.

InvGate empowers organizations by providing the tools to deliver seamless service across departments, from IT to Facilities.

When and the way did you initially change into eager about computer science?

My interest in computer science dates back to my early childhood. I used to be all the time fascinated by electronic devices, often finding myself exploring and trying to grasp how they worked. As I grew older, this curiosity led me to coding. I still remember the fun I had writing my first programs. From that moment on, there was little doubt in my mind that I desired to pursue a profession in computer science.

You might be currently leading R&D initiatives and implementing novel generative AI applications. Are you able to discuss a few of your work?

Absolutely. In our R&D department, we tackle complex problems that will be difficult to represent and solve efficiently. Our work is not confined to generative AI applications, however the recent advancements on this field have created a wealth of opportunities we’re keen to take advantage of.

Considered one of our primary objectives at InvGate has all the time been to optimize the usability of our software. We do that by monitoring the way it’s used, identifying bottlenecks, and diligently working towards removing them. One such bottleneck we have encountered often is expounded to the understanding and utilization of natural language. This was a very difficult issue to handle without the usage of Large Language Models (LLMs).

Nonetheless, with the recent emergence of cost-effective LLMs, we have been capable of streamline these use cases. Our capabilities now include providing writing recommendations, robotically drafting knowledge base articles, and summarizing extensive pieces of text, amongst many other language-based features.

At InvGate, your team applies a method that is named “agnostic AI”. Could you define what this implies and why it will be important?

Agnostic AI is fundamentally about flexibility and flexibility. Essentially, it’s about not committing to a single AI model or provider. As an alternative, we aim to maintain our options open, leveraging the most effective each AI provider offers, while avoiding the chance of being locked into one system.

You’ll be able to consider it like this: should we use OpenAI’s GPT, Google’s Gemini, or Meta’s Llama-2 for our generative AI features? Should we go for a pay-as-you-go cloud deployment, a managed instance, or a self-hosted deployment? These aren’t trivial decisions, they usually may even change over time as recent models are released and recent providers enter the market.

The Agnostic AI approach ensures that our system is all the time able to adapt. Our implementation has three key components: an interface, a router, and the AI models themselves. The interface abstracts away the implementation details of the AI system, making it easier for other parts of our software to interact with it. The router decides where to send each request based on various aspects reminiscent of the sort of request and the capabilities of the available AI models. Finally, the models perform the actual AI tasks, which can require custom data pre-processing and result formatting processes.

Are you able to describe the methodological points that guide your decision-making process when choosing probably the most suitable AI models and providers for specific tasks?

For every recent feature we develop, we start by creating an evaluation benchmark. This benchmark is designed to evaluate the efficiency of various AI models in solving the duty at hand. But we do not just give attention to performance, we also consider the speed and value of every model. This provides us a holistic view of every model’s value, allowing us to decide on probably the most cost-effective option for routing requests.

Nonetheless, our process doesn’t end there. Within the fast-evolving field of AI, recent models are continually being released and existing ones are usually updated. So, each time a brand new or updated model becomes available, we rerun our evaluation benchmark. This lets us compare the performance of the brand new or updated model with that of our current selection. If a brand new model outperforms the present one, we then update our router module to reflect this transformation.

What are among the challenges of seamlessly switching between various AI models and providers?

Seamlessly switching between various AI models and providers indeed presents a set of unique challenges.

Firstly, each AI provider requires inputs formatted in specific ways, and the AI models can react in another way to the identical requests. This implies we want to optimize individually for every model, which will be quite complex given the variability of options.

Secondly, AI models have different capabilities. For instance, some models can generate output in JSON format, a feature that proves useful in lots of our implementations. Others can process large amounts of text, enabling us to make use of a more comprehensive context for some tasks. Managing these capabilities to maximise the potential of every model is a vital a part of our work.

Finally, we want to be sure that AI-generated responses are secure to make use of. Generative AI models can sometimes produce “hallucinations”, or generate responses which might be false, out of context, and even potentially harmful. To mitigate this, we implement rigorous post-processing sanitization filters to detect and filter out inappropriate responses.

How is the interface designed inside your agnostic AI system to make sure it effectively abstracts the complexities of the underlying AI technologies for user-friendly interactions?

The design of our interface is a collaborative effort between R&D and the engineering teams. We work on a feature-by-feature basis, defining the necessities and available data for every feature. Then, we design an API that seamlessly integrates with the product, implementing it in our internal AI-Service. This permits the engineering teams to give attention to the business logic, while our AI-Service handles the complexities of coping with different AI providers.

This process doesn’t depend on cutting-edge research, but as a substitute on the appliance of proven software engineering practices.

Considering global operations, how does InvGate handle the challenge of regional availability and compliance with local data regulations?

Ensuring regional availability and compliance with local data regulations is an important a part of our operations at InvGate. We fastidiously select AI providers that can’t only operate at scale, but in addition uphold top security standards and comply with regional regulations.

As an example, we only consider providers that adhere to regulations reminiscent of the General Data Protection Regulation (GDPR) within the EU. This ensures that we will safely deploy our services in several regions, with the arrogance that we’re operating inside the local legal framework.

Major cloud providers reminiscent of AWS, Azure, and Google Cloud satisfy these requirements and offer a broad range of AI functionalities, making them suitable partners for our global operations. Moreover, we repeatedly monitor changes in local data regulations to make sure ongoing compliance, adjusting our practices as needed.

How has InvGate’s approach to developing IT solutions evolved during the last decade, particularly with the mixing of Generative AI?

During the last decade, InvGate’s approach to developing IT solutions has evolved significantly. We have expanded our feature base with advanced capabilities like automated workflows, device discovery, and Configuration Management Database (CMDB). These features have greatly simplified IT operations for our users.

Recently, we have began integrating GenAI into our products. This has been made possible because of the recent advancements in LLM providers, who’ve began offering cost-effective solutions. The mixing of GenAI has allowed us to reinforce our products with AI-powered support, making our solutions more efficient and user-friendly.

While it’s still early days, we predict that AI will change into a ubiquitous tool in IT operations. As such, we plan to proceed evolving our products by further integrating AI technologies.

Are you able to explain how the generative AI inside the AI Hub enhances the speed and quality of responses to common IT incidents?

The generative AI inside our AI Hub significantly enhances each the speed and quality of responses to common IT incidents. It does this through a multi-step process:

Initial Contact: When a user encounters an issue, he or she will be able to open a chat with our AI-powered Virtual Agent (VA) and describe the problem. The VA autonomously searches through the corporate’s Knowledge Base (KB) and a public database of IT troubleshooting guides, providing guidance in a conversational manner. This often resolves the issue quickly and efficiently.

Ticket Creation: If the problem is more complex, the VA can create a ticket, robotically extracting relevant information from the conversation.

Ticket Task: The system assigns the ticket to a support agent based on the ticket’s category, priority, and the agent’s experience with similar issues.

Agent Interaction: The agent can contact the user for extra information or to notify them that the problem has been resolved. The interaction is enhanced with AI, providing writing recommendations to enhance communication.

Escalation: If the problem requires escalation, automatic summarization features help managers quickly understand the issue.

Postmortem Evaluation: After the ticket is closed, the AI performs a root cause evaluation, aiding in postmortem evaluation and reports. The agent also can use the AI to draft a knowledge base article, facilitating the resolution of comparable issues in the long run.

While we have already implemented most of those features, we’re continually working on further enhancements and enhancements.

With upcoming features just like the smarter MS Teams Virtual Agent, what are the expected enhancements in conversational support experiences?

One promising path forward is to increase the conversational experience right into a “copilot”, not only able to replying to questions and taking easy actions, but in addition taking more complex actions on behalf of the users. This may very well be useful to enhance users’ self-service capabilities, in addition to to offer additional powerful tools to agents. Eventually, these powerful conversational interfaces will make AI an ubiquitous companion.

LEAVE A REPLY

Please enter your comment!
Please enter your name here