Quick Answer: Honasa Consumer, the parent company of Mamaearth, has appointed Vipul Maheshwari as chief product and technology officer (CPTO). He will lead product, technology, engineering and analytics, and his main job is to advance the company's use of artificial intelligence, including agentic AI. Agentic AI refers to systems that carry out multi-step tasks with human oversight.
Honasa Consumer just gave one executive a much bigger remit. Maheshwari, who had been executive vice president for product and data, now runs the company's product, technology, engineering and analytics functions under a single CPTO title. Most coverage stops at the job title. The more useful question is what this consolidation says about how a direct-to-consumer (D2C) beauty and personal care brand plans to operate, and what you can learn from it if you run a product or data team.
What Honasa actually changed
According to ETtech's report, the company informed the stock exchanges of the appointment. Maheshwari was already a member of the senior management team, so this is an elevation rather than an outside hire. That matters. Internal promotions usually signal continuity of strategy, not a pivot.
The scope is what changed. Under the new title, he oversees four functions that many consumer companies still keep in separate silos: product, technology, engineering and analytics. Honasa has also said it is building AI tools and workflows to simplify processes, automate tasks and speed up decisions. Putting one leader over all four makes it easier to move an AI idea from a data model to a shipped workflow without handoffs.
Why the CPTO title matters more than the org chart
Here's the counter-intuitive part. Most readers assume that adding "technology" to a product title is cosmetic, a way to signal that the company is serious about AI. In practice, the reporting line is what decides whether AI work ships. When analytics sits under finance or marketing, a model might find a pricing opportunity that nobody in product is empowered to act on. When engineering sits in a separate silo, the model's output waits in a backlog.
Bringing those functions under one owner removes that wait. It doesn't guarantee better results, though. A combined role can also concentrate too much responsibility in one person, and it can blur the line between building AI and governing it. Good CPTO setups usually keep an independent check on model quality, often through a data governance or risk function that doesn't report to the same leader.
What agentic AI actually means in practice
Agentic AI is the term people use for AI systems that pursue a goal across several steps. A chatbot answers one question. An agent reads a request, decides which tools to call, executes the steps, checks the results and escalates when something looks wrong. The source describes it as systems that carry out multi-step tasks with human oversight, and that oversight part is where most real-world projects succeed or fail.
Consider a simplified example. A brand notices a stockout risk on a bestselling serum. A basic AI tool flags the risk on a dashboard, and a planner decides what to do. An agentic workflow goes further: it checks inventory levels, drafts a replenishment order, compares supplier lead times and proposes a reorder for approval. The human still signs off, but the planner now reviews a finished recommendation instead of building it from scratch.
The gap between those two versions is where the value sits. Maheshwari himself framed it this way: AI and agentic systems move businesses "from tools that provide information to intelligent systems that can execute workflows, connect decisions and drive outcomes." That's a useful test you can apply to any AI pilot in your own organisation. If the tool only informs, it's a dashboard. If it executes a workflow with a checkpoint, it's an agent.
Why his background fits the role
Maheshwari has worked at Honasa for six years and helped build its data and product capabilities, according to the report. He holds a master's degree in mathematics and computer science from the Indian Institute of Technology (IIT) Delhi. Before joining Honasa, he worked at Delhivery, the logistics company, and at Global Analytics India.
That mix matters for agentic AI specifically. Logistics teaches you that operational data is messy and that small errors compound across a network. Analytics work teaches you to distrust a model that looks good on average and fails on the edge cases. Those are the instincts you want in someone who will decide which workflows an AI agent can execute without review.
CEO and cofounder Varun Alagh put the rationale plainly, saying technology has helped the company grow and AI offers a chance to rethink how it builds and runs the business. He credited Maheshwari's "strong foundation in data and product, along with his deep understanding of our business." Internal knowledge is a real advantage here. A new hire would need months to learn which processes are brittle.
The business context behind the move
Honasa's results give the appointment some financial backing. Its net profit more than doubled to Rs 90 crore in April-June, up from Rs 41 crore a year earlier, its highest quarterly profit. For the July-September quarter, the company expects consolidated net sales value (NSV) growth in the early thirties percent year-on-year.
Be careful with that profit jump. Doubling profit from a low base is easier than sustaining it, and a single strong quarter doesn't prove a structural change. The more reliable signal is the guidance for NSV growth, because it tells you management expects demand to hold up. Whether AI contributes to that growth is a question the next several quarters will answer.
The risks you should watch
Agentic AI in a consumer brand has specific failure modes. The first is pricing and promotion errors. An agent that adjusts discounts across thousands of SKUs can erode margin in hours if its guardrails are weak. The second is regulatory claims. Personal care products in India operate under advertising rules, and the Advertising Standards Council of India (ASCI) polices misleading claims. A generative tool that drafts product copy without a compliance check can create liability fast.
The third risk is quieter: oversight fatigue. When an agent produces 200 recommendations a day, reviewers start approving them without reading closely. That's when human oversight becomes a formality. The fix is to limit each agent's scope, log its decisions and sample its outputs regularly, not just rely on the approval step existing.
Here's where most teams get it wrong. They treat AI governance as a legal document written once. It works better as an operating routine, with named owners and measurable thresholds.
How the old and new roles compare
| Area | Before the appointment | After the appointment |
|---|---|---|
| Title | Executive vice president, product and data | Chief product and technology officer (CPTO) |
| Functions covered | Product and data | Product, technology, engineering and analytics |
| Stated priority | Product and data capabilities | Advancing AI, including agentic AI, across the business |
| Status | Member of senior management | Appointed via stock exchange filing |
Frequently Asked Questions
What is agentic AI?
Agentic AI describes systems that plan and complete multi-step tasks toward a goal, using tools and data along the way. Unlike a chatbot that answers one prompt, an agent can check inventory, draft an order and wait for approval. Human oversight is a standard part of responsible deployments, and it's what separates a useful agent from an unsupervised one.
What does a chief product and technology officer do?
A chief product and technology officer sets the direction for how products are built and how technology supports them. In Honasa's case, the role covers product, technology, engineering and analytics. Combining these under one leader shortens the path from an idea to a shipped workflow, although it also concentrates accountability in one person.
Why does Honasa Consumer need a CPTO for AI?
Honasa needs one leader who controls both the data and the product decisions, because AI workflows fail when those are split across teams. The company has said it is building AI tools to simplify processes, automate tasks and speed up decisions. A single owner can make trade-offs between speed and risk without waiting for committee approval.
How can a brand evaluate an agentic AI pilot?
Start with one bounded workflow, such as reorder planning, and define what the agent may do without approval. Measure accuracy against a human baseline, log every decision and review a sample each week. Expand scope only after the error rate stays low for several cycles. This approach keeps the risk contained while you learn where the agent helps.
Keep the takeaway simple
The most useful lesson from this appointment is structural: AI delivers results when the people who own data, product and engineering share one set of priorities. Try mapping where your own data, product and engineering work currently hand off to each other, then pick one handoff to shorten this week and note what changes. Pass this to anyone on your team who's wrestling with how to make Honasa Consumer-style CPTO thinking work for their own business.