Alumni Ventures: Talkingly AI – Sanchali Sharma, PGP 2020
In this conversation, Sanchali Sharma (PGP 2020) reflects on a career shaped by curiosity, technology, and a commitment to building for inclusion. From leading AI-first products at global organisations to co-founding Talkingly AI, she shares insights on entrepreneurship, leadership under uncertainty, and what it takes to build meaningful technology in a rapidly evolving AI landscape. Grounded in experience and honest self-reflection, her journey offers a thoughtful perspective on innovation that is both ambitious and human-centred.
To begin with, could you tell us a little about yourself and what you’re currently working on?
I am a seasoned AI product leader who has worked on multiple AI first consumer and enterprise products. I am an engineer by training and then got fascinated by strategic thinking, product disruption, entrepreneurship and knack for 0->1. My mission of life is to democratise technology and AI to enable and empower everybody. As part of my professional journey, I have worked with Microsoft building various enterprise and consumer products such as exchange, teams, bing, which helped me solve user problems around communication and information leveraging disruption innovations happening in machine learning, deep learning, and AI. I also spent on the other extreme of edge AI in the field of production, manufacturing and supply chain at Emerson. After that I co-founded two startups where I conceptualised and built semi-autonomous agentic solutions for sales AI, retail AI, recommendation systems, consumer AI and implementing sentient web at talkingly AI.
You co-founded Talkingly AI. What sparked the idea behind the company, and what problem were you most keen to solve when you started?
I have always been fascinated by the enablement of the internet and mobile. Widespread and accessible internet enabled commerce and communication across the earth, realising the ideals of vasudaiva kutumbkam. However, we still see that there were barriers due to accessibility, language, technology awareness, age. I saw my parents struggling to leverage the benefits of mobile because they find it difficult to navigate. At the same time, I also realised that speaking is the skill we humans acquire before even writing.
Talkingly AI works towards the idea of a more “sentient” web. How do you explain this vision to someone outside the AI ecosystem?
The idea of sentient web is to make websites intelligent so that any human can interact with them in a way that is most natural to them, speaking, viewing, screen readers, braille interface or say through VR devices. This has massive implications about enabling access to information and content across the barriers of language and accessibility. Further because we are leveraging generative conversations which cater to the user needs, we are not just making websites accessible, we are transforming websites into agencies of task completion. A powerful and subtle aspect is that now a website owner doesn’t need to think about how to layout the information to cater to different needs of the users. They have to cater to providing the information and the AI will autonomously leverage the provided content to best suit the user needs.
Building Talkingly AI has involved leading product, engineering, and research. What has this journey taught you about leadership and decision-making?
Leading product, engineering, and research at Talkingly AI has been less about hierarchy and more about orchestrating conviction under uncertainty. I often think of Peter Drucker’s insight: “The most important thing in communication is hearing what isn’t said.” In practice, this meant that leadership wasn’t issuing directives but it was listening closely to engineers, researchers, and customers, spotting friction points before they became crises, and nudging the team towards clarity.
Decision-making here has been a lesson in marrying evidence with intuition. As Winston Churchill quipped, “However beautiful the strategy, you should occasionally look at the results.” Every launch, pivot, or failed experiment reminded me that data without context is just noise and gut instinct without feedback is recklessness. The trick was and is balancing the two in real world scenarios.
From raising pre-seed and seed rounds to being recognised among the Top 10 startups in North England and pitching at London Tech Week, what have been some defining moments in this journey?
Raising pre-seed and seed rounds, being named among the Top 10 startups in North England, and pitching at London Tech Week weren’t just milestones—they were proof that Talkingly AI was moving from experiment to impact.
1. Idea Validation from Real voices, real validation
The first real breakthrough came when a customer used our audio-first agent to resolve queries on their website and said, “this just saved me hours of back-and-forth.” That moment proved our concept of conversational AI that listens, understands, and acts across the world wide web was genuinely solving a problem, not just ticking boxes.
2. Pre-Seed Funding – Investors betting on us, not just the tech
Closing the pre-seed round wasn’t about the code, it was investors believing we could create value in an emerging, complex AI space. It reinforced that clarity of vision, conviction, and a deep understanding of the problem are as important as the tech itself.
3. Staying Close to Customers – Product roadmap guided by real need
We stayed glued to customer feedback: which queries stumped Talkingly, which interactions delighted users, which features genuinely reduced friction.
4. Recognition – Validation from the ecosystem
Being named a Top 10 startup confirmed that Talkingly AI’s approach was not just clever, but relevant.
5. Pitching on Big Stages – Storytelling that converts curiosity into confidence
Pitching at London Tech Week forced us to articulate not just how Talkingly works, but why it matters. Demonstrating live how our AI handled complex queries convinced audiences and investors that this wasn’t just another bot, but a platform transforming how businesses interact with their customers and make real revenue for them
London Tech Week is where we met additional partners who got really interested in the technology behind the UI.
Every entrepreneurial journey has moments of uncertainty. Was there a phase that truly tested you, and how did you navigate it?
Talkingly AI tested me in ways no case study ever could. We were working with audio-first AI models at a time when the technology was still emerging, the market was undefined, and even sophisticated investors struggled to form a clear view of ROI because the stack itself was complex and unfamiliar. There were multiple false starts, features that didn’t land, narratives that didn’t resonate, and moments where conviction had to be rebuilt from first principles.
We stayed relentlessly close to our customers’ voices, letting their feedback shape every iteration. That real-world grounding not only kept us on the right path but also helped convince investors that our approach was solving a genuine problem, turning uncertainty into clarity and clarity into product roadmap. It was a demanding phase, but it fundamentally shaped how I make decisions under ambiguity today.
Looking back, how do you see your time at IIM Bangalore shaping the way you build, think, and lead today?
IIM Bangalore taught me to be comfortable with uncertainty long before startups made it fashionable. The case-studies conducted in classrooms, especially in management & economics, had a habit of dismantling neat theories the moment incentives, information asymmetry, or regulation entered the picture. Professors didn’t just teach how markets should work, they showed us how businesses actually behave when money, power, human vices and risks are involved.
That grounding shaped my entrepreneurial instinct: take informed risks, understand second-order effects, and never confuse a good model with the real world. It’s a mindset I still lean on when building products and leading teams across the UK and US.
As an alumna building globally, what does being part of the IIM Bangalore alumni community mean to you now?
Today, the IIM Bangalore alumni community feels like a trusted inner circle rather than a formal network. It’s a group I can turn to for unfiltered advice, sharp debates, and occasional reality checks , regardless of geography or time zone. There’s a shared language shaped on campus: practical, rigorous, and refreshingly low on pretence, which makes the community both grounding and genuinely useful as I build globally.
If you were to meet your younger self—perhaps on campus—what advice would you offer her today?
I’d tell her to stop optimising for validation and start optimising for learning velocity. The applause on campus fades faster than you think, but curiosity and the ability to build things compound quietly over time. Also, don’t wait to feel “ready” , that moment rarely arrives, and if it does, it’s usually late.
Sanchali’s journey underscores a recurring theme: progress rarely comes from certainty or validation, but from sustained learning, listening closely, and building with purpose. Whether navigating emerging technologies, leading teams through ambiguity, or reimagining how humans interact with the web, her reflections offer a nuanced view of leadership in practice. Her advice to her younger self—optimising for learning velocity and beginning before feeling ready—serves as a fitting takeaway for anyone building at the edge of change.
