The end of the ticket is only the beginning

On June 12, Atlassian brought together Indian technology and business leaders from industries spanning FinTech, retail, automotive software, fashion, lending, food delivery, industrial safety equipment, and more. The occasion was the roundtable, ‘The Ticketless Future Is Here. Are You Ready?’. The crux was whether organisations are willing to look honestly at how they work before they hand decisions to a machine.

The roundtable featured Vivek Iyer, Head of Product for IT Ops and Assets at Atlassian's Jira Service Management; Aravind Raghunathan, Head of AI at Murugappa Group; Piyush Garg, Director of Engineering at Jubilant FoodWorks; Sourav Dasgupta, CIO at Allcargo Global; Abhinav Tiwari, AVP (Engineering) at the Open Network for Digital Commerce or ONDC; Thiyaga B, AVP Engineering at CaratLane; Shakir Wani, Head of Data Science at Aditya Birla Fashion and Retail; Tanmoy Deb, DGM of Software at Motherson Innovations; Arjun Marwaha, Head of Technology at YabX; and Amit Bhatia, Group CIO at KARAM.
The 95% problem
The first candid admission came from Aravind Raghunathan of the Murugappa Group. His company had deployed an AI coaching tool for operational agents, a system that surfaced insights, recommended next actions, and nudged people toward better decisions. Although it worked technically, almost nobody used it.


When the team investigated, it found that the organisation had been running for years on informal knowledge and unspoken assumption, with standards labelled 'good' in one department meaning something different in another, and documentation that existed on paper but not in practice. "We thought the process was always followed," Raghunathan shared. "We thought there was a proper repository being maintained by the team, but that never happened on the ground."

AI, in this telling, was the diagnostic that made visible what had always been broken.

Abhinav Tiwari from ONDC identified the most stubborn version of this problem: unclear ownership. Fragmented knowledge can be addressed with documentation, and inconsistent processes can be improved. But when an engineer leaves, what walks out with them isn't just their code, but the reasoning behind every architectural decision they made. "What will be missed is why he chose that architecture over another. That's where the ownership problem comes in,” he said.

Piyush Garg described a parallel experience at Jubilant FoodWorks, where AI now monitors store performance in real time on high-stakes sale days and generates dynamic interventions (discount triggers, menu changes) when a store looks likely to miss its target. The system functions, but the complication is deciding what the system is allowed to do.

"Wherever three things are involved: the customer, your revenue, and your brand perception, you need to define guardrails," Garg said. "AI has to work within those boundaries only." The business, it turned out, had never needed to be that explicit about those boundaries before. AI made it necessary.

Sourav Dasgupta came at the same problem from a different angle. Rather than deploy AI on a high-visibility project, the company has been making it democratic, giving every employee basic AI capability, starting with clearer emails and sharper presentations. One employee's standout presentation set off a wave of imitation across the organisation. "It builds healthy competition and it builds a culture. And it tells you, ‘we can do it’,” he underlined.

Vivek Iyer connected all of this to what Atlassian has spent two decades building: structured institutional memory, in Confluence documents and Jira histories, that can serve as context for AI reasoning. The argument is that if an organisation hasn't kept streamlined, consistent records of what happened and why, the AI will fill in the gaps with guesses.

"If there's nothing in the document, AI will hallucinate much more because it will make mistakes. It will give you incorrect decisions,” he underlined. “"Context is going to be very critical. If you improve context, you will improve AI quality. But once you solve context, the next question is to apply it to make AI repeatable and accurate."

In essence, the problem isn't the model, but everything upstream of it.

Right answer, wrong experience
Some of the sharpest material came from the consumer-facing end of the table. Thiyaga B of CaratLane described the difficulty of deploying AI in jewellery retail, where a single purchase might cost ₹2-3 lakh and trust takes years to build. AI produces three reactions, he said: awe when it gets something exactly right; unease when it feels like it knows too much; and real damage when it gets something wrong. "We can't bring a solution which is almost there," he said. "In the trust part, it has to be 100%."

Shakir Wani described a parallel gap at Aditya Birla Fashion & Retail, where an AI set generator producing outfit combinations performed well online but met some resistance on the shop floor, from sales staff with decades of floor instinct who weren't ready to trust a system they hadn't themselves figured out. The accuracy numbers told one story; the five combinations that flopped told another. "The team will mention those five flop sets and say, ‘That's why we have reservations’,” he pointed out.

The system wasn't wrong, but it was operating in an environment where trust is earned individually, not statistically, and where the people who needed to use it had no history of relying on something they hadn't themselves figured out.

Reflecting on both accounts, Atlassian’s Iyer made the point that AI quality is still in its infancy, and the honest response is not to lower standards but to raise the rigour of evaluation. An analogy he reached for was cloud computing a couple of decades ago, a technology that was met with similar institutional suspicion until the track record made resistance feel more costly than adoption.

The end of the ticket
When the conversation arrived at its central question of “if service tickets disappeared, what would be missed most?”, the answers were more nuanced than the provocation invited.

Abhinav Tiwari said ownership would vanish first. Tickets exist to assign someone to a problem. Remove the ticket, and you remove the clearest signal in most organisations for who is responsible. Piyush Garg said visibility: even if AI resolves a problem before a human notices it, leadership still needs to know what happened, what action was taken, and why.

Vivek Iyer reframed the stakes. The ticket is just the current unit of measuring work done, and replacing it doesn't mean abandoning accountability, it means moving it downstream. An agent can log what it fixed even when no human noticed the problem. The ticketless future isn't the absence of a record, but the record being written by the system rather than the person.

Meanwhile, Tanmoy Deb steered the conversation to the demands of his industry. In automotive software, where a product must function reliably for years, the stakes of silent resolution are considerably high. The concern is whether there's any trace of the decision if something goes wrong two years later.

Arjun Marwaha from YabX, which operates in the heavily-regulated world of lending-as-a-service, went further. In fintech, an audit trail is the answer to every question a regulator, customer, or court might eventually ask. "The audit trail should tell me ‘How did it reach this decision? What were the policies? Was it within the boundary?’ I, as a reviewer, should be able to trace it completely. And those audit logs should be immutable,” he emphasised.

Who watches the watchman?
One of the sharpest technical moments came when Raghunathan was asked what concerned him more, an AI agent making a poor decision, or one that overlooked something while appearing to function normally.

The answer, for Raghunathan, was the second by a considerable margin. A bad decision is visible. If a payment gateway opens the wrong bank's net banking portal, the customer can't pay, the transaction fails, and a monitoring alert fires. Now imagine the platform processing thousands of orders without error, except that somewhere in the chain, the amount printed on the physical invoice doesn't match what the customer paid.

"With AI, there are new kinds of nuances we need to monitor: toxicity, model drift, context overflow, prompt injection. It's no longer just CPU 100%, memory 100%, disk 100%,” he said.

Amit Bhatia of KARAM, which exports safety equipment to 150 countries, echoed Raghunathan. AI that overlooks something is worse than AI that decides wrongly, he said, because an overlooked gap suggests the model hasn't learnt that dimension at all. A bad decision can at least be traced and corrected with better data. A blind spot requires recognising that the blind spot exists, which is hard to see from inside.

The takeaway
The ticketless future isn't primarily a technology problem. Those at the roundtable who were furthest along were the ones who’d spent time understanding how decisions get made within their walls, as distinct from how the organisation chart says they should be made.

The second thing that emerged is that structured, maintained, and honest institutional memory (read: context) is the real constraint. Get the quality right, and auditability becomes possible. Without context, everything else is theoretical.

The third was that the ticketless world may already be further along than most organisations realise. The question isn't whether AI can manage work without human instruction, but whether the humans responsible for those systems know what's happening, and whether they've built the structures to find out when something goes wrong.