It’s hard to escape the pull of LinkedIn, especially if you’re a millennial or younger who wants to network online in a professional setting. The whole world is, of course, on the likes of Facebook, WhatsApp, Instagram and Snapchat, but LinkedIn is inherently different according to Malai Lakshmanan – a key person influencing its trajectory. “LinkedIn has been a professional network and we’ll continue to be one,” he says.
“If you look at the kind of content people share – especially with all the advancements in tech – we’re seeing strong engagement from members who come to learn what’s happening in their field, how it impacts them, and how it will impact their jobs and careers,” says Malai Lakshmanan, Senior Director and Head of India Engineering at LinkedIn, in an exclusive interview.
When I ask him if LinkedIn is turning into Facebook – the irritating parts that perhaps take away from the network’s professional feel – Lakshmanan sets the record straight. “We do get this question once in a while. To a certain extent, it depends on the quality of your feed – who you’re connected to and what topics you’re interested in. It’s a reflection of that,” he emphasizes.
“Some people I know say LinkedIn is the only social media network they’re on,” Lakshmanan explains further. “And because it’s their only social platform, they sometimes share something more personal with their community here. But for many of them, it’s also because they don’t want the noise of other networks and find LinkedIn to be high-value.”
Of course, equating LinkedIn with Facebook isn’t the only misconception about the professional social network, in Malai Lakshmanan’s perspective. According to him, people wrongly attribute LinkedIn’s value to the size of their personal network on the platform.
“One misconception I often see – especially among entry-level talent – is the belief that they can’t get value from LinkedIn unless they already have a big network,” he says. He recommends simple things like just creating a profile, exploring, expressing interests, following the right topics and people, and engaging with relevant content as indicators of a strong start on LinkedIn.
Also read: LinkedIn’s new AI features can write your cover letter, review resume & more
“You don’t need a large network from day one to begin seeing value. Over time, your network and opportunities naturally grow as you engage with the platform in ways aligned to your interests,” he explains.
And while LinkedIn’s perception in the public eye is often defined by what shows up on your feed, Lakshmanan wants people to understand how deeply India is embedded into the platform’s global engineering engine. LinkedIn Engineering, he explains, spans multiple layers – the flagship application experience where members interact with the platform, a vast data layer where member information is treated like a “crown jewel,” and a complex infrastructure layer that keeps everything running reliably at scale.
“India has a meaningful presence across these layers. On the customer-facing side, a lot of our Talent Solutions work is run out of India,” says Malai Lakshmanan, including several features and systems that millions of users interact with daily on LinkedIn – often without realising that teams in India are behind them.
“For example, the Easy Apply experience on LinkedIn is run by a team in India,” he mentions. Lakshmanan also points out the work of making LinkedIn Recruiter connect neatly with the broader ecosystem is also driven out of India. “Teams in India power ingestion so jobs from customer platforms appear on LinkedIn. We also run Talent Insights out of India.”
Every action on LinkedIn – a connection request, a comment, a like, a follow – generates an event. Those events aren’t just analytics trivia; they become the raw material for improving feeds, refining recommendations, and measuring product success. “That entire tracking infrastructure is actually run out of India,” he says. In other words, the systems that validate and route these events into LinkedIn’s data pipelines are India-owned – and foundational to everything from business insights to data science experiments.
Experimentation, too, is a huge part of LinkedIn’s operating rhythm, he adds. Whether it’s a subtle UI tweak or a major product change, experiments run continuously across the platform to understand what members respond to. “We do a lot of experimentation across the platform – everything from UI variations to broader product changes. There’s significant infrastructure to support experimentation, and that is also run out of India. These are critical to how the company globally tracks what members like and how the business is doing,” highlights Malai Lakshmanan.
Perhaps the most intriguing infrastructure story Lakshmanan shares is how an existing system built for something else became relevant to GenAI almost overnight.
LinkedIn runs a universal schema registry that helps its data pipelines move faster by centralising how data attributes are defined. “Passing both data and metadata can be heavy. The schema registry centralizes those definitions so only the data needs to flow, improving performance,” Lakshmanan points out. Recently, LinkedIn faced a new need of centralising GenAI prompts used across engineering, both for efficiency and compliance.
Rather than build a new system from scratch, the India team repurposed the universal schema registry to act as a GenAI prompt registry – essentially a central store of prompts that can be monitored and governed, aligning with LinkedIn’s “members-first” approach to data access. “A really small tweak,” Lakshmanan calls it, but one that now powers a company-wide capability to keep up with AI-enabled workflows.
LinkedIn also receives large volumes of member and customer support queries. Lakshmanan says, “We’re working on models that can interact conversationally (including voice), guide people to the right information, check satisfaction, and resolve issues faster – reducing dependency on traditional ticketing workflows.”
Also read: LinkedIn will use your profile data to train its AI model: Here’s how you can stop it
The point, he suggests, isn’t just automation. It’s improving the quality and speed of resolution for members, while allowing support teams to scale without proportional headcount growth.
On the product side, Malai Lakshmanan frames GenAI as the entry point – and agentic workflows as the real evolution. “A recent example is LinkedIn Hiring Assistant, designed to make recruiters’ work more effective. Recruiters deal with heavy workloads – thousands of applications, screening, outreach drafts, and more. Hiring Assistant can help filter candidates, draft messages, and support the recruiting workflow through conversational interactions,” he says. The recruiter stays in control, but the machine does the grunt work.
If there’s one disruption Lakshmanan is watching for in 2026, it’s not “GenAI” in the generic sense. “The rise of agentic experiences is particularly exciting,” he says, “we may start seeing not just individual AI tools, but how they interact with each other to serve a purpose – where the combined value becomes greater than the sum of individual parts.”
He acknowledges the growing backlash, of how AI gets things wrong, returns feel overpromised, and expectations are running ahead of reality. “But it’s still nascent,” Lakshmanan emphasizes. “Our worlds are complex, and expecting immediate perfection is unrealistic. The interesting shift will be when these agentic systems connect and coordinate – within enterprises and across the industry.”
Also read: OpenAI thinks it can rival LinkedIn in AI hiring: Here’s why
For students and young professionals, Lakshmanan reinforces the idea that “skills are the new currency,” pointing to relevance in areas like LLMs, prompt engineering, AI literacy and code reviews. But he also stresses on something even more important, “Human skills are becoming even more important – communication, getting your ideas across, negotiation, empathy, and understanding different viewpoints,” as AI frees up time from heads-down execution.
Towards the very end of our interview, LinkedIn’s Malai Lakshmanan flips the script by sharing that he was reverse-mentored by interns – not as a cute leadership gimmick, but because he believes today’s entry-level talent is arriving with AI-native thought process.
“We’ve seen a shift. Earlier, engineers would think first and then ask AI. Now, people are learning to think more AI-native – bringing AI into the workflow differently. So adaptability is required on both sides. Entry-level talent can’t follow old playbooks, and experienced professionals can’t assume incremental change is enough. This wave is a marked shift, and all of us need to rethink how we work, what skills we build, and how we lead,” concludes Malai Lakshmanan.
