Customizing AI for a Competitive Edge
What if AI could personalize the experiences your customers have with your brand? Perhaps you’ve started to use generative AI yourself. Did you summarize a long document, or build a customized chatbot? As AI becomes easier to access, organizations are increasingly customizing AI to create unique apps and experiences that differentiate them in the marketplace.
The NBA is a great example, redefining fanship with AI-powered personalization, delivering tailored stats and highlights. The city of Buenos Aires transformed urban living by using ‘Boti,’ an AI chatbot that handles questions for residents about driver’s license renewals, subway schedules, and tourism—managing over two million monthly queries.
These organizations are bending AI to their vision, pushing the limits of what’s possible.
I’m excited to share findings from a new MIT Technology Review Insights report, which explores how businesses use AI customization to maintain a competitive market advantage: DIY GenAI: Customizing generative AI for unique value. The report highlights the motivations, methods, and challenges technology leaders face when tailoring AI models to create value for their businesses.
While customizing AI isn’t new, its potential is expanding. According to the MIT report, while boosting efficiency is a top motivation for customizing generative AI models, creating unique solutions, better user satisfaction, and greater innovation and creativity are equal motivators.
- Improved efficiency is often the first clear benefit. As organizations gain experience, the learning curve flattens, and we’ll likely see other benefits as companies focus on customizing AI to increase revenue.
Specializing with Agents
When choosing AI models, half of the leaders surveyed in the MIT report are prioritizing agentic and multi-agent capabilities, in addition to multimodality (56%), flexible payment options (53%), and performance improvements (63%). AI agents perform tasks and make decisions, with great utility for:
- Autonomous problem-solving in areas like data entry and retrieval for clinical operations in Healthcare
- Supplier coordination and maintenance tracking in manufacturing
- Enhancing inventory and store operations in Retail
Agents have the potential to transform markets by providing unique solutions beyond automating processes that humans find dull.
Atomicwork, an ITSM and ESM platform, is an example of this. They offer specialized AI agents that integrate into the flow of work, providing support without multiple tools or complex integrations. According to Atomicwork, one of their customers achieved a 65% deflection rate (the percentage of issues resolved without human intervention) within six months.
Good Data Equals Good AI
The potential of AI customization is immense, but it also has its challenges. Data integrity is the biggest barrier. Half of participants in the MIT report cited data privacy and security (52%) and data quality/preparation (49%) as obstacles. However, generative AI presents innovative ways for companies to interact with and use their data in unique solutions.
Critical to empowering data-driven AI is an intelligent data platform that unifies, governs, secures data, and seamlessly integrates with AI building tools. Microsoft Fabric is the fastest-growing analytics product within the company’s history. Microsoft is also seeing AI-driven data growth, and Fabric removes the obstacle of data integrity.
RAG is the Customization Starting Point
One of the simplest and most effective customization methods is retrieval-augmented generation (RAG). Two-thirds of those surveyed in the MIT report are implementing or exploring RAG. Grounding an AI model in an organization’s specified data is unique and capable of providing a specialized experience. In practice, RAG is often used with fine-tuning (54%) and prompt engineering (46%) to create very specialized models.
Dentsu, an advertising / PR firm, developed custom techniques for generating media channel insights for clients. By integrating a customized RAG framework and an agentic decision layer, Dentsu reports about 95% accuracy in retrieving relevant data and insights. This AI-powered approach is central to campaign strategies and optimizing marketing budget allocation for their clients.
Empowering Development Teams
Model features and capabilities, along with developer tools, evolve rapidly. Therefore, empowering teams with the right tools is crucial for successful AI. With the pace of new model capabilities, automating model evaluation is critical. According to the MIT report, 54% of companies use manual evaluation methods, and 26% are either just starting to apply automated methods or are doing so consistently.
The report notes that playgrounds and prompt development features are also widely used to facilitate collaboration between AI engineers and app developers while customizing models. Full lifecycle model evaluation is built into Azure AI Foundry, to continuously evaluate model capabilities, optimize performance, test safety, and keep pace with advancements.
Looking Ahead with Azure AI
AI has high utility for creating services and experiences that can set you apart in the marketplace. Models are continually advancing and specializing by task and industry. Today, there are more than 1,800 models in the Azure AI Foundry catalog–and they are evolving just as quickly as the tools and methods to build with them. Agents delivering new customer service experiences will reshape customer service. AI customization will become standard practice for building with AI.
What’s that unique experience for your business? What’s the next special project that you want to do for your clients or customers? How do you want to empower your employees? You’ll find everything you need to bend the curve of innovation with Azure AI Foundry.
No matter where you are in retooling your organization to operationalize AI, I encourage you to read the MIT report. The team spoke with technology leaders about creating unique value by customizing generative AI. Download the MIT report.
