Explore how Swapnil Jain, Co-Founder & CEO of Observe.AI, envisions agentic AI transforming contact centers, enhancing CX, and driving measurable ROI.
Swapnil, could you start by sharing a bit about your professional background and what inspired you to co-found Observe.AI?
Absolutely. I started my career as an engineer who was deeply curious about how technology could solve real-world problems. Before Observe.AI, I worked on building scalable systems at Twitter. But during my tenure in India with Twitter, I saw firsthand how broken the customer service experience could be. Agents were overwhelmed, customers frustrated, and almost no agents were used to improve performance despite all the data flowing through contact centers. That was the spark – could we use AI to make every conversation count? That vision led to Observe.AI.
What are the key challenges enterprises face when building and operationalizing truly autonomous, agentic AI solutions?
- Data readiness – Most enterprises don’t have structured, high-quality data to train AI effectively.
- Trust and governance – Deploying autonomous systems requires transparency, explainability, and strong safeguards, especially in regulated industries.
- Integration into workflows – Even the smartest AI is useless if it can’t integrate into how people and systems work.
How can organizations distinguish between “AI-washed” products and truly autonomous AI solutions with advanced reasoning, empathy, and brand alignment?
- Demonstrated outcomes – Is the solution delivering measurable automation, ROI, and customer experience gains at scale? Do they have existing customers like you in terms of size and scale? Ask to speak to customers and understand what it took to make the project successful.
- Depth over demos – Can it go beyond scripts and simple logic to reason, personalize, and align with your brand voice? Ask for a short proof of concept that you deploy in real-world enterprise environments with measurable automation and outcomes.
- Enterprise-readiness – Does it offer audit trails, compliance, and the ability to iterate and improve in production? Can the AI reflect your tone, values, and policies in how it responds, escalates, and acts?
In your experience, why are contact centers emerging as the ideal proving ground for next-generation AI applications like GenAI and VoiceAI?
They have high-volume, high-variance data, real-time decision-making needs and a direct impact on customer experience and business outcomes. They’re also typically under pressure to reduce costs while improving quality, making them fertile ground for transformation.
What measurable benefits have you observed or expect from deploying VoiceAI in contact centers, particularly regarding cost savings and ROI?
I’ve seen customers reduce operational costs by 50% by automating routine interactions. Others have achieved new growth with VoiceAI features like 24/7 chatbots that speak 20 languages, which was impossible earlier due to headcount issues. And for some, ROI isn’t simply about automation; it is about augmenting agents to perform better and extracting insights that drive the business forward.
What essential AI infrastructure components are required to move beyond human-in-the-loop copilots toward fully autonomous decision-making agents in production environments?
A conversation-aware runtime that can reason, make decisions, and take meaningful action, not just generate responses. It leverages real-time and historical data for context and includes a trust layer with explainability, fallback logic, governance controls, and deep integration with CRMs, ticketing systems, and knowledge bases.
How does the integration of agentic AI impact organizational workflows and employee roles, especially in complex enterprise settings?
It’s a shift from task-based work to outcome-oriented roles. For example; agents become exception handlers and escalation specialists rather than script followers. QA teams focus on insights instead of scorecards. This shift also unlocks new roles, like AI trainers and conversation designers, which didn’t exist before.
What role does empathy and brand alignment play in designing autonomous AI agents, and how can these qualities be effectively embedded?
They’re critical. AI agents are now customer-facing, and they represent your brand. Embedding empathy requires AI agents with emotion detection, sentiment analysis, and real-world interactions. Brand alignment comes from tone, vocabulary, and escalation strategies, all of which must be customizable and continuously improved through control frameworks and feedback loops.
Looking ahead, what trends or innovations do you foresee in the deployment of agentic AI that enterprises should prepare for now?
I see three main innovations emerging, sooner than enterprises think:
- Multimodal agents that operate across voice, chat, email and help agents in real-time.
- Proactive AI that initiates conversations, not just reacts.
- Federated AI systems where humans and machines co-manage tasks in real time.
Enterprises must start preparing their data infrastructure, change management strategies, and talent models immediately.
What advice would you give to enterprise leaders who are just beginning their journey toward adopting agentic AI and autonomous decision-making systems?
Start small but think big.
Pick high-volume, low-risk use cases to test and learn. Build internal champions across operations, IT, and compliance. And do not wait for perfection – iterative deployment with strong governance will get you faster to ROI and transformation.