AIONOS is a JV between InterGlobe Enterprises (IndiGo Airlines) and the Assago Group. No wonder therefore when Abhinandan Jain, President – CX & Global Head of Marketing, AIONOS exclusively informs Rajneesh De, Group Editor, APAC Media and CXO Media about its deployment of a 24×7 intelligent concierge across voice, chat, web, and mobile, handling flight status, gate information, and baggage tracking instantly, in English, Hindi, and regional dialects with Adani Airports.
Could you share an overview of the company’s core offerings and how it is helping businesses accelerate their digital transformation journeys?
AIONOS was built on a conviction the market is only now catching up to: the bottleneck in enterprise AI is not the model, it is the orchestration. In most organizations, data, analytics, automation, and human expertise sit in silos that cap their real impact. As a joint venture between InterGlobe Enterprises and the Assago Group, we deliberately chose not to compete in the race to build foundation models. We compete on accountability, on connecting insight to decision-to action inside live enterprise systems.
UniWeave is our agentic CX layer. It orchestrates a network of specialized agents toward a defined outcome i.e. a booking confirmed, a claim settled, a churn risk reversed with certified connectors into the systems of record that run an enterprise: Amadeus and Sabre in travel, core banking and BSS/OSS in telecom, SAP and Blue Yonder in logistics. UniStack is the control plane that routes each request to the most cost-effective model, enforces guardrails, and records the reasoning behind every material decision. Our Cloud Analogy acquisition brings deep Salesforce Agentforce engineering in-house.
The acceleration comes from collapsing the value chain. A typical enterprise stitches together a model provider, an integrator, and an operations partner, then absorbs the friction between them. We deliver all three as one accountable operating model. Our June 2026 alliance with Black Box extends across more than 35 countries, pairing their digital infrastructure with our applied AI so enterprises get a single owner of the journey from physical layer to intelligence layer.
How are AIONOS offerings different from other generative AI and automation platforms in the market today? Can you share a real-world example?
Most platforms ship chatbots, and a chatbot optimizes for a reply. UniWeave optimizes for an outcome. Every conversation is tied to an outcome graph, not a static decision tree, with multi-agent planning that decomposes a task, selects tools, and carries memory across sessions and channels. That is a different architecture from the generative-AI-as-a-feature approach still dominating the market, and it is where the frontier is moving. Cisco’s data shows over half of support interactions will run on agentic AI by mid-2026, and Gartner expects agentic systems to autonomously resolve 80% of common service issues by 2029. We built for that endpoint, not the current one.
A concrete example from aviation: the conventional passenger help desk means queues, repetitive disruption handling, and agents consumed by routine queries. Our deployment with Adani Airports puts a 24×7 intelligent concierge across voice, chat, web, and mobile, handling flight status, gate information, and baggage tracking instantly, in English, Hindi, and regional dialects. Routine volume resolves autonomously, freeing human agents for the complex, high-emotion moments where judgment decides the outcome.
The same pattern holds globally in telecom, banking, and logistics. And the next step is already visible in our roadmap: agents that act before the customer does. When a flight cancels or a shipment stalls, the system rebooks, applies the credit, and notifies the customer before they pick up the phone. That shift from reactive to proactive is the real differentiator, and most of the market has not built the orchestration to do it.
How do you address concerns around data privacy, security, and responsible AI usage while deploying AI-powered customer experience solutions at scale?
This is the real ceiling on enterprise AI, not capability, and it is consistent across every market we operate in. As autonomy rises, the binding question stops being “can the AI do it” and becomes “can we control and account for what it did.” We engineered for that from the start rather than treating compliance as a final checkbox.
Every agent action in UniWeave runs through a layered guardrail engine policy, privacy, brand, regulatory, and financial with deterministic checks that sit alongside the model rather than after it. That distinction matters. A guardrail that fires after the model has acted is theater. We constrain what an agent can say, do, and commit to before it acts. The platform is SOC 2 compliant, runs end-to-end encryption, and offers data residency options, which is decisive across the EU, India, and the Gulf, where rules are tightening at different speeds.
The differentiator is auditability. UniStack records the reasoning pathway behind every major decision, so when an agent makes a material choice, the enterprise can reconstruct exactly why. Accountability is not a promise that the AI behaves; it is a verifiable trail proving how it behaved. As enterprises move from agents that assist to agents that act independently, that audit layer becomes the precondition for scaling at all, not a feature you add later.
What industries are currently witnessing the fastest adoption of AI-driven CX solutions, and what lessons can other sectors learn from their success?
The fastest movers share a profile: high interaction volume, thin service margins, and low tolerance for delay. Travel, aviation, telecom, banking, and logistics lead globally for that reason. In India, BFSI, retail, healthcare, and industrials are projected to drive roughly 60% of the net new AI value toward a $500 billion opportunity by FY2026, and the same vertical pattern holds across the Gulf, Southeast Asia, and Europe.
The first lesson for everyone else is to start where volume and friction intersect, not where the technology looks most impressive. Prove value on the high-frequency, high-frustration journey first. The second is that leaders redesigned the journey rather than deflecting customers off it. The ones who used AI to push people away from humans saw satisfaction fall; the ones who used it to resolve faster and route hard cases to better-equipped humans improved cost and experience together.
The third is the one most sectors underestimate. McKinsey finds only about 23% of organizations are actually scaling agentic AI while 39% are still experimenting, and most deployments are stuck in one or two functions. The gap is not ambition, it is readiness clean data, knowledge hygiene, and the governance to let agents act. The technology is increasingly the easy part. Winners prepare data, process, and people in parallel and treat data foundation as a precondition, not a phase-two cleanup.
When you deploy AI for a massive client, how do you decide which tasks the AI should handle completely on its own, and when a real human employee needs to step in?
We decide along two axes: complexity and consequence. High-frequency, rules-based tasks with low downside balance inquiries, status updates, routine rebooking are owned end to end by the agent. That is where autonomy delivers the speed and consistency customers prefer. The moment a task carries high emotional stakes, real ambiguity, irreversible consequences, or regulatory sensitivity, a human takes it.
But the real engineering is the handoff, not the line. A transition where the customer repeats everything destroys the value of both sides. UniWeave escalates proactively when sentiment drops below a threshold, honors a customer-initiated transfer instantly, and passes full context to the agent so the human starts informed. We run agent-assist in the other direction too, with real-time suggestions, knowledge retrieval, and sentiment coaching.
I would push back on the assumption that this is mainly about removing headcount. The mature model running across leading enterprises in 2026 is “AI assist” and “AI act” operating side by side, with human leaders supervising autonomous workflows in real time and adjusting thresholds based on outcomes. The boundary is dynamic. As an agent proves reliability under continuous monitoring, it earns more autonomy. It is a graduated trust model, and the human role shifts from handling volume to governing the system, a more demanding job, not a disappearing one.
What trends are you observing among enterprises when it comes to cloud adoption, AI implementation, and customer experience transformation?
Enterprises have moved decisively past experimentation. The question three years ago was whether to adopt; now it is how fast and how to govern. India sits at the front of this curve, not just consuming AI but building and exporting it, yet the shift is universal across mature and emerging markets alike.
On cloud, the posture has matured from lift-and-shift to deliberate optimization for cost, residency, and the right hybrid mix. On implementation, the constraint has moved from technology to data readiness and talent, and the enterprises pulling ahead invested early in clean, accessible data. On CX specifically, the most consequential dynamic I see is scale combined with linguistic and cultural diversity — acute in India but equally real across the Gulf, Africa, and Southeast Asia. Serving customers across dozens of languages and across deeply digital and not-yet-digital populations is a problem agentic AI is uniquely suited to solve.
Looking ahead, the real shift is architectural: single-threaded automation is giving way to multi-agent systems where agents pass context, share memory, and coordinate across functions and eventually across organizations, agents transacting with other companies’ agents. By 2026, AI in CX is no longer a differentiator; it is the baseline. The differentiator becomes who can orchestrate autonomous journeys safely while keeping humans central to the moments that matter. My caution is that many enterprises are over-indexing on scale and speed under competitive pressure without asking whether the system delivers lasting value. The winners, anywhere in the world, will treat AI as a solution to a genuine operational problem and let growth follow the value rather than using it to mask the absence of any.
