Chatbots Take Over the Front Desk in Travel Booking

Continuing my scan of how India Inc is using AI, today I’m looking at the online travel portals. EaseMyTrip lists several AI initiatives, which I’ve summed up below. The immediate focus — for this company and for most customer-facing firms — is the interaction layer for booking queries, details and data. 

The goal is simple: shorten resolution time, cut manpower and costs, and make the process feel smoother. In practice, most of them are pushing more of the detailed back-and-forth onto the customer. The chatbot becomes the first stop, and you’re left navigating a mixed bag of responses.

I’ve used the Emirates and IndiGo chatbots myself, for details on booking, PNR, additions and so on. Both have their strengths. It’s nice not to hear the endless “your call is important to us” loop. But the moment you ask anything even slightly off-track, the bot’s rigid logic shows up- a "non linear" query still suggest you talk to the travel consultant. 

Still, there’s a lot of AI being deployed across the sector — including on our side. One of our expertise, our Jetmetaphy chatbot can now handle several layers of queries with reasonable depth.

Human in the loop is critical in AI

Why the Human in the Loop remains critical was highlighted by Sudipta Roy, MD and CEO of L&T Finance,  during the Q&A of the Investors Digital Day, November 6 2025, at the Jio Conference centre ( summarised and paraphrased): 

Project Cyclops delivers full automation for smaller-ticket loans — such as two-wheelers with average sizes of ₹1.1–1.2 lakh — where the risk is low and the model’s confidence is high.

But for larger value loans, L&T is more careful. 

For large-ticket loans (₹25–50 lakh and above), the machine’s decision is treated only as first-level underwriting. A human underwriter must still corroborate the output. The reason is simple:
AI systems can have bugs, and early versions of Cyclops produced anomalous results that were only caught because humans reviewed the files.

For high-value cases — ₹60 lakh to ₹1 crore — the human check remains non-negotiable until the system has built 12–18 months of performance history that proves its reliability.

In the interim, the company's Helios co-pilot acts as an assistive layer, supporting underwriters and ensuring the machine’s decisions are validated before approval.

From Dealer Offers to Risk Outcomes: What Cyclops Reveals About AI in Credit

As part of my ongoing coverage of how Indian NBFCs are using AI in real, operational settings, I looked at remarks from Asheesh Goel, Chief Executive of Farmer Finance at L&T Finance, during the company’s Investors Digital Day. He outlined the company’s AI underwriting and dealer-offer optimisation system, Cyclops, and shared early results from its rollout.

The goal here isn’t to spotlight one company. It’s to understand how Indian lenders are applying AI at scale, what’s working, and what other financial institutions can learn when they plan their own deployments. India Inc is moving fast on this front, and many of these improvements are replicable across the sector.

Using the transcript and my own prompt, I summarised Cyclops’ performance in the table below (Dealer coverage, STP gains, LTV shifts, GNS outcomes, and more). The numbers aren’t meant to sell a success story. They show something more practical: AI is starting to move core credit, risk, and productivity metrics inside Indian NBFCs, not just peripheral customer-facing tasks.

This kind of transparency helps the wider industry understand what modern AI systems can actually deliver, what timelines look like, and how business teams adapt to machine-assisted underwriting. It also sets a benchmark for other lenders evaluating similar projects.

Cyclops is just one example, but it highlights broader lessons:

1. AI value is clearest when tied to a business bottleneck
   Here, the bottleneck was pricing, underwriting, and dealer engagement — not some generic “digital transformation.”

2. Full-funnel improvement beats isolated metrics
   Offer quality, STP, LTV, ticket size, and GNS all moved in the right direction. That’s what sustainable AI adoption looks like.

3. Execution speed separates leaders from laggards
   A five-month rollout is not typical in BFSI. It shows the impact of product-like execution rather than project-like execution.

4. Risk-adjusted gains are the strongest signal
   A drop in GNS and non-starter rates suggests that the model isn’t just automating decisions; it is refining them.

5. This isn’t unique or proprietary — it’s replicable
   None of the gains require exotic research. They require clean data, disciplined deployment, and a business team that is willing to trust the system.

I am tracking other applications of AI- will cover going ahead. 

AI at Scale: L&T Finance’s Enterprise Transformation

The shift toward AI and digital tools is obvious in another NBFC: L&T Finance.

In its November 6, 2025 Investors Digital Day presentation, the company noted that in March 2021 only 14% of servicing queries were handled through digital channels. Call centres handled 35% and branches managed 51%. By September 2025 the picture had flipped. Digital channels handled 93% of all servicing queries, while call centres were down to 2% and branches to 5%.

That’s the digital shift on its own. The AI focus is even more striking. In the company’s Digital Day meeting on November 6, 2025, there were 66 direct mentions of the term “AI,” along with many more references to related projects. I pulled this count from an AI scan I ran using my own prompt.

source: investor digital day, November 6 2025 transcript 

The document is worth reading in full- its been summarised above for easy reference. Check out the version I created using NotebookLM here

One section shows a sample interaction with KAI, their conversational servicing bot. It’s able to get customers to complete payments about half the time. That’s a direct 50% cut in human effort, cost, and turnaround time.

We see the same pattern at Jetmetaphy. We build and run agentic AI systems for vendor workflows, QA, invoice tracking, and other routine processes. May not glamorous, but simply practical, efficient and profitable! 

Bajaj Finserv Shows What Scaled AI Looks Like in India

Building on my earlier post about Bajaj Finserv and the ₹5,300 crore it expects to generate from robocalls in FY26, here’s a bit more detail from the analyst presentation on November 25, 2026 (available on their website). If you’re interested in how AI is actually being implemented and used at scale, make sure you read the Analyst Presentation dated December 5, 2025.

AI Implementation Summary 

    source: Bajaj Finserv Analyst presentation (read this document for more) 

46% of all interactions by AI for non-voice BOT ; 9% for voice BOT and 4% for real time agent assistance. 

What did all this deliver? 

• For Bajaj General Insurance, 96% digital policy issuance, and 1 out of every 2 customers serviced entirely by AI chat & voice bots 
• ₹7.9 crore annual savings through AI-led service automation 
• 110 million messages exchanged annually via bots 
• 29% of servicing in Indian languages thanks to multilingual capabilities 
• English : 71% , Hindi 15% , Telugu 6% , Marathi 5% , Tamil 3% 

source : Bajaj Finserv analyst presentation 

The company clearly is fully committed to AI now. It’s not a side project anymore. They’re using it at real scale, and it’s becoming a core part of their competitive moat. Plenty of companies talk about doing this, but this is one of the rare cases where we can actually see the numbers and understand the scale of AI adoption in an Indian industry setting.

How AI Voice Bots Drove ₹5,300 Crore in Loans—and Why It Matters

There was an interesting piece of news earlier this week. Bajaj Finserv, one of India's largest NBFC ( non banking finance company) expects to generate Rs 5300crore from voice bots over FY26 ( April 2025- March 2026). 

source: The Times of India 

Bajaj Finserv pushing out ₹5,300 crore in loans through voice bots says a few things.

First, they can now write far more loans with far fewer people. That’s automation doing exactly what it’s meant to do. If their prediction is right and digital channels take over by 2030, the thousands of young workers spending their days on outbound calls will see those jobs disappear. You could argue these jobs were already draining and rejection-heavy, but the shift is still huge.

Second, the number shows that millions of consumers are comfortable talking to a machine about money. Either they’re tech-forward, they trust the system, or they just want to get a quick answer and move on. Any of those reasons is a big cultural shift. A loan conversation used to feel personal; now it’s routine enough to outsource to a bot.

Third, it raises a question. Why would anyone discuss a personal loan of up to ₹50 lakh with a robot? Most people would expect at least one human conversation for something that size. My guess is the bot’s job is simple: filter for people who press “1,” pass them to a human, and hand over a qualified lead. After that, expect persistent follow-ups. Once you’re “in the system,” it’s hard to get out.

Fourth, the calls themselves point to a larger issue: how did lenders get all these phone numbers? Everyone in India knows the answer. Numbers are scraped from all over-  websites, entry logbooks, apartment registers—wherever they can be found. Privacy rarely enters the picture, and AI only speeds up this harvesting.

Fifth, AI makes the whole pipeline smoother. It can stitch together data from several sources, build a profile, generate a pitch, and have a bot deliver it. It can guess when you might need money and contact you at the right moment. That’s efficient, but also unsettling.

The bigger picture: AI is turning lending into an extremely data-heavy industry, maybe even on par with old government databases. And it shows how fast AI is reshaping the way people deal with companies, with each other, and with institutions.

The ₹5,300 crore figure is basically proof that the shift isn't theoretical anymore. It's already happening at scale. And given the scale of this business, do not expect robocalls to stop anytime soon. 

How AI Is Rapidly Rewriting the Rules of Economics

AI is reshaping economics as a study and profession at high speed. Economics has always been a mix of human emotion, logic, irrationality, fear, and hope. The variables keep shifting as new information and circumstances roll in. It is exactly the kind of environment AI is built for, with its expanding memory base and ability to spot patterns as they form.

AI already fits naturally across the discipline. Econometric modeling is an obvious starting point. Take inflation. Instead of relying on a handful of indicators to estimate CPI or WPI, models can now draw on thousands of variables. With the right training, AI can map inflation trends with far more precision and update its assessments in near real time, giving planners and policymakers a faster, deeper view and maybe a better window to act.

The same applies to forecasting and modeling. AI can test scenarios almost instantly, pulling in global, regional, and local factors. It can estimate GDP with finer detail, track how fuel prices ripple through hundreds of industries, and simulate policy impacts across a wide spread of sectors.

Behavioral economics also gains. Instead of static surveys and linear extrapolation, AI can read context as it unfolds and produce a richer picture of sentiment and behavior.

And of course, as I keep noting, AI can read text at scale and extract sentiment across sources. That alone is gold for political strategists.

This is only the start. AI excels when relationships multiply beyond what the human mind can track. Even so, the human in the loop stays essential. The best computer is still the one in our heads. AI just helps it work smarter.