Grateful to join Shripati Acharya and the Prime Ventures Podcast to discuss Uniphore’s journey and what it really takes to bring agentic AI into production at the world’s largest businesses. We covered a few themes that keep coming up with CIOs and business leaders I speak with: what industrializing AI looks like beyond pilots, why a horizontal enterprise platform (data, models, security, agents) is essential for scaling responsibly, and the challenge of making AI agents more deterministic through architecture, guardrails, and workflow design. I also shared why small, fine-tuned models can be the right approach for many business use cases, balancing accuracy, control, and cost at scale. At the center of this is the Uniphore Business AI Cloud: an end-to-end platform that helps businesses bring their data together, fine-tune models, and build and deploy AI agents with the security and sovereignty large organizations require. And, it’s increasingly enabled by a strong ecosystem of strategic partners across compute, data, and cloud, working together to help customers move from experimentation to measurable outcomes. I’m confident the next chapter of Business AI will be defined by disciplined execution and real impact at scale. Thank you Shripati for a thoughtful conversation. Let’s do it again soon. #BusinessAI
The Holy Grail of Business AI: Uniphore's $2.5 Billion Playbook with Umesh Sachdev
https://www.youtube.com/
Ibrahim Bafakih
2mo
The winners won’t be the ones who try AI they’ll be the ones who operationalize it.
One of the best podcasts to understand AI adoption at scale. Simple, straight to the point. My takeaways: 1. Enterprise decisions need to be thought as horizontal platform enabling where business users make their own AI companion (think framework, tooling, security, platform etc from that lens) 2. Agentic Adoption might not be very wide yet, but it is deep at places, and it is creating significant impact where its deep (revenue leakage example) 3. Best learning from talk - agentic AI in 4 layers at Uniphore Data layer - First level, agents to find data fit for purpose Knowledge layer - Fine-tuned SLMs (fast, cost saving) Model layer - Guardrail, explainability, reasoning Agentic layer - Singular task, Singular agent, BPM framework, multiple agent orchestrations Thanks for amazing talk - Umesh Sachdev Shripati Acharya