fbpx

Gen AI and ROI: Bridging the gap from hype to reality in enterprise adoption

Gen AI and ROI: Bridging the gap from hype to reality in enterprise adoption 2560 1493 Kane Simms

In the world of technology, few innovations have sparked as much interest as generative AI (gen AI). This emerging technology promises to transform how businesses operate and serve their customers. However, while the potential of gen AI is immense, many organisations face significant hurdles in their journey from experimentation to large-scale deployment. During a recent webinar with Vaibhav Bansal, Vice President at Everest Group and Murali Raghuram. SVP Product, Kore.ai, we explored the key challenges, solutions, and what the future might hold for enterprises adopting Gen AI.

The challenge of finding a clear use case

One of the biggest hurdles businesses face when adopting gen AI is determining how to use it. Many companies struggle to identify which processes or functions gen AI could meaningfully enhance. While gen AI can be applied across various industries and tasks, pinpointing a specific, valuable use case can be difficult. Murali noted that this is a common theme: “There have been a lot of companies I have been speaking with recently who are working on things like AI policies. And there’s the risk, compliance, and legal teams. Everyone is seemingly not entirely confident about moving forward with generative AI.”​

Beyond use case identification, companies often grapple with the cost of scaling gen AI across thousands of users. This is particularly challenging as organisations move from pilot projects to full-scale implementation. The complexity and lack of clarity around the return on investment (ROI) also hold many businesses back. As Vaibhav pointed out, new use cases – those that haven’t existed before and which could fully harness the power of AI – are still relatively uncommon. Many companies are still in the early stages of using AI to improve existing services and processes, rather than creating new ones​.

Organisational and regulatory barriers

Another significant barrier is organisational resistance. Enterprises are often held back by internal concerns from risk, compliance, and legal teams. Many are hesitant to move forward due to data security, privacy, and regulatory compliance uncertainties. This hesitancy can be attributed to a lack of understanding and clear frameworks around how generative AI will impact operations and governance.

Moreover, businesses must comply with regulations varying by region, particularly regarding data privacy and security. Just look at LinkedIn’s recent terms update. It opted everyone who’s outside the EU into sharing their data for use in training its models, but couldn’t do that by default to those inside the EU.

This is crucial for industries like healthcare, where AI applications in drug discovery or diagnostics could drastically shorten timelines but are still bound by stringent regulatory oversight​.

Solutions to overcome barriers

Despite these challenges, solutions are emerging to help organisations navigate the complexities of generative AI adoption. One critical strategy is the creation of “guardrails,” which refer to the frameworks and boundaries set to ensure that AI models perform within acceptable parameters. These include defining thresholds for risk scores and ensuring that AI-generated outputs do not compromise sensitive information or violate compliance rules. For example, businesses can apply guardrails directly to their AI agents, setting limits on what the models can do based on their risk assessments​. While this isn’t a 100% reliable solution today, performance is improving constantly.

Furthermore, platforms like GALE, a gen AI application builder from Kore.ai, make it easier for companies to experiment and implement AI without needing extensive coding expertise. The no-code platform allows businesses to build, test, and deploy AI-driven applications rapidly. GALE’s integration with other enterprise systems ensures businesses can seamlessly embed gen AI into their existing workflows, avoiding disruptions​​.

Another critical solution is the fine-tuning of models. Instead of relying solely on large-scale models, companies can create smaller, more specific models that better serve their needs, especially in edge cases where performance and latency are crucial​. This ability to distil custom models from larger datasets provides a scalable and cost-effective way for businesses to tailor gen AI to their operations.

Measurable results: ROI and metrics

The ROI of gen AI remains a hot topic. Many enterprises still find it challenging to quantify the impact of their AI initiatives, particularly because most early applications have focused on improving existing processes rather than creating new value. However, experts at the webinar suggested that businesses should focus on a few key metrics when evaluating the success of gen AI. Depending on the specific use case, these could include customer experience, operational efficiency, and time-to-market improvements​.

Vaibhav also emphasised the importance of balancing costs with potential benefits, noting that organisations must account for the infrastructure costs and the long-term expenses of maintaining and scaling AI solutions. For many, cloud-based solutions provide a more affordable and scalable option compared to on-premises infrastructure​​.

The future of gen AI in business

As gen AI continues to evolve, its potential to revolutionise industries is becoming clearer. From enhancing customer service with AI-driven agents to accelerating drug discovery, the technology’s applications are broadening. However, businesses must navigate the challenges of organisational support, cost, and regulatory compliance to unlock AI’s full potential.

The key takeaway from the webinar is that gen AI adoption is still in its early stages, with many companies experimenting but few at full production scale. As more use cases emerge and solutions like GALE make AI more accessible, the technology’s widespread adoption seems inevitable. However, businesses must be strategic to succeed, prioritising use cases that offer measurable value and building the necessary guardrails to ensure compliance and security.

Gen AI offers transformative potential, but success will require careful planning, investment, and a willingness to experiment. Those who can balance these elements will likely be the leaders in this next wave of digital innovation.

Missed the live webinar, no worries! Check out the recording here.

    Share via
    Copy link
    Powered by Social Snap