Shopping Cart

Close

No products in the cart.

Inside the gold rush to operationalise AI across enterprises

​[[{“value”:”

In the last few years, the business landscape has witnessed a seismic shift, thanks to the introduction of generative AI technologies like ChatGPT. GenAI is capable of composing original text, images, computer code and other content, and has gone from an experimental curiosity to one of the biggest business disruptors of our generation.

Generative AI: A Golden Opportunity

The frenzy of interest in genAI is akin to a digital gold rush, as businesses race to mine and organise data into 24 carat opportunities. While the fields sparkle with promise, businesses must sift through opportunities carefully, separating AI hype from practical reality, and finding ways to move forward while staying one step ahead of the competition. EXL stands at the forefront of this transformative wave, assisting insurance companies, banks and financial utilities to get a handle on their data so they can implement sustainable, AI-driven customer support technologies.

Overcoming Implementation Challenges

While the spotlight has only just started shining on genAI, the underlying tech has existed for a while, and many pioneering companies have already paved the way for successful enterprise-grade implementations. In fact, the team has been implementing generative and conversational AI in complex professional services for nearly a decade. This experience has given us some unique insights into the implementation of these technologies, including common roadblocks, challenges and real-world timelines.

While the excitement around generative AI is palpable, its implementation is not without concerns, especially in regulated industries. We advise our clients that building a proof of concept using genAI takes about 2-3 weeks. From there, actual implementation requires robust data governance, data migration, and a breaking down of organisational silos. In most cases we are seeing these tools first launched and trained internally, rather than being immediately customer-facing. Think of genAI implementation as a marathon rather than a sprint. 

Popular Use Cases for GenAI 

Generative AI is transforming businesses, particularly in job functions that involve processing large volumes of information to synthesise insights. This technology is rapidly advancing and continually improving at an astronomical rate, but this doesn’t mean humans will be forced out of the loop. It will, however, change what jobs we have and the way that we work.

Off-the-shelf genAI engines from heavy hitters such as OpenAI, Google and Amazon are certainly powerful, but the real power lies in developing and implementing company-wide commercial genAI initiatives. Businesses that want to get ahead in the AI arms race need to prioritise data management, organisation and accessibility, so they can use comprehensive reference data to train these systems for specific applications.

We are currently seeing the fastest and most impactful commercial applications of genAI in the following four areas:

Agent Support in Customer Service

These AI-driven tools utilise past customer interactions, payment records, transaction details and investment history to provide customer service representatives with tailored response scripts. This includes applications like verifying insurance eligibility for prescriptions and collection processes.

The implementation of generative AI in our pilot program for collection processes and agent queries proved successful in providing real-time nudges, procedural guidance and call summarisation. As a result, we anticipate achieving a remarkable 45-50% reduction in average handling time and a notable 50-60% increase in agents’ speed to competency.

Contract Analysis and Drafting

In the finance, insurance and legal industries, genAI is being used to extract crucial details from contracts and legal documents to address inconsistencies, identify risks and even draft legal text.

This enables claims adjusters to quickly access a comprehensive summary of claim information through a user-friendly interface powered by open-source large language models like Pythia, Vicuna and Flan-T5. This application aims to improve the efficiency and accuracy of the information retrieval process, resulting in a 25% reduction in the time agents spend on addressing customer queries.

Audit and Compliance Revolution 

Businesses in finance and accounting are adopting generative AI for comprehensive analysis of compliance documents. This approach is replacing the traditional method of manual auditing based on random samples.

The broking application undergoes a thorough document review to ensure all required fields are filled and check for data completeness, security and compliance before manual data entry takes place. Additionally, our partnered solution offers privacy LLM training and penetration testing and automatically generates compliance documentation to assist in meeting privacy requirements.

Additional Use Cases

Generative AI has been utilised to address various use cases, including code generation, hyper-personalised marketing messages, customer service, and personalised financial dialogue.

Operationalising GenAI: Data Governance and Human-in-the-Loop 

Across all these applications, two elements stand out as pivotal for transforming technological innovations into viable, commercial-grade solutions: robust data governance and human expertise. This is what differentiates genuine, commercial-grade solutions from mere technological novelties. 

In complex, highly regulated industries like insurance and financial services, simply applying an off-the-shelf AI chatbot to a front-end interface is not an option. To deliver real value and minimise risk, these applications should not only have governed access to and understand a spectrum of customer data, they should also be trained on the subtleties of specific use cases to ensure accuracy and adherence to regulation, security and privacy.

Early leaders are quickly realising that solid data quality, architecture and governance are not only beneficial, but essential prerequisites for a successful genAI application. They are also recognising the equally important role of humans-in-the-loop. From programming and training large language models, to seamlessly integrating this technology into existing workflows and interpreting specialised subject matter, gen AI will continue to require human intermediation. The true potential of AI lies not in replacing human capabilities, but in augmenting our strengths and mitigating our weaknesses.

Where to Next? Embracing the Future with Generative AI

GenAI revolution has only just begun, so it’s essential businesses look beyond the initial hype, embracing the opportunity to leverage AI in real ways to enhance their workflow intelligence and customer and employee experience. Successful operationalisation of genAI is not just about adopting the technology, but integrating it strategically and ensuring it adds value and drives sustainable growth. Those who embrace this technology wisely will undoubtedly emerge as the leaders of tomorrow.

Keep up to date with our stories on LinkedInTwitterFacebook and Instagram.

“}]] 

Leave a Reply