Buying insurance should feel like securing peace of mind. Instead, it feels like walking into a maze. For instance, when you open a health insurance site, you are instantly overwhelmed with jargon like room rent capping, co-payments, and waiting periods for pre-existing conditions. The words blur into confusion and mistrust. To add to the chaos, there are hundreds of insurance products in the market. At this point, most people either give up or buy something reluctantly, with a knot of anxiety.
This broken experience is what Ruchir Kanakia, founder of Bangalore-based OneAssure, set out to fix. Launched in 2020, OneAssure is a full-stack insurance platform aiming to fundamentally re-engineer the way health and term insurance are discovered, purchased, and serviced in India. But their approach isn’t what you might expect.
Instead of burning cash to sell insurance directly to consumers (D2C), they decided to build the OneAssure AI Advisor to empower the industry’s most overlooked yet critical players: the humble insurance distributors or agents. OneAssure AI advisor acts as a copilot, helping agents explain complex policies simply, match customers to the right plan, and deliver clarity where the industry usually breeds confusion. It turns agents from salespeople into trusted consultants.
Today, their bet on AI is paying off. Over 100 active insurance distributors rely on the OneAssure platform, processing transactions across more than a dozen of India’s top insurers. The result is an unheard-of 87-90% month-on-month retention rate. By augmenting human trust with AI, OneAssure is solving a deep-seated pain point for the industry’s frontline and fundamentally changing how insurance is sold and served.
Ruchir shares his product journey in a conversation with Team ProdWrks.
Solving Insurance with First Principles
Ruchir Kanakia didn’t come from the insurance industry. He came from tech. An engineer by training, he was part of the founding team at cloud telephony pioneer Exotel and held leadership roles at Instamojo and Cuemath.
He first dipped his toes into the insurance world in 2019, initially building software for insurers. He found that the system was riddled with misaligned incentives and technological neglect, where the customer was the ultimate casualty.
"The individual agent, who drives the majority of distribution, is averse to using technology. The insurance companies don't want to change. Everyone has their own agenda, and in all of this, the customer is lost."
"When you start your online research to buy insurance, you see a lot of taglines saying 'these are the five best products.’ They're good as taglines, but the true answer is unique to an individual. In retail insurance, you have more than 300 products for any need. Even if you apply the 80/20 rule, you're still left with 60 products to choose from. How do you narrow that down?"
‘OneAssure AI advisor’ empowers insurance agents to help buyers
For most tech founders aiming to disrupt a legacy industry, the playbook is simple: go direct-to-consumer (D2C), build a slick app, and out-market the incumbents. Ruchir and his team at OneAssure made a deliberate, counter-intuitive choice.
They built an AI product to aid insurance distributors and agents. They didn’t try to sell insurance to the consumers directly.
Ruchir’s reasoning reveals a deep understanding of both market dynamics and human psychology. He was solving for the mistrust that people had when buying insurance.
"You would trust someone whom you know more than you trust me or any platform today. You trust your uncle who lives in your society, or you trust your friend's mother who's selling to you, because they're less likely to cheat you."
Instead of trying to replace this existing trust network, OneAssure decided to supercharge it. Their GTM strategy was to arm these trusted local advisors (your Chartered Accountants, wealth managers, the neighbourhood insurance agents) with technology that would make them exponentially more effective, transparent, and efficient. It was a B2B2C play, building a scalable product by solving the problems of the people who already owned the customer relationship.
The second reason why Ruchir chose the B2B2C route was purely business logic.
He says, "I didn't start selling directly to consumers as the CAC (Customer Acquisition Cost) is too high for scaling in B2C. I also need to solve for being a sustainable business."
Decoding the OneAssure Platform
To understand the impact of OneAssure’s platform, you first need to appreciate the daily grind of an independent insurance distributor. They work with insurance companies that have clunky, unreliable portals. A simple request for a premium comparison for a client means logging into three or four different apps, manually inputting data, screenshotting results, and compiling them in a messy PDF or Excel sheet to share on WhatsApp. It’s a soul-crushing, inefficient process.
OneAssure’s platform replaces this chaos with a single, lightning-fast dashboard. On the OneAssure platform, a distributor can input a client’s details once and get quotes from over a dozen insurers in seconds. They can select three or four plans, click “Compare,” and generate a clean, shareable link with a detailed comparison that is easy to understand.
“I can just send it to you right now on WhatsApp,” Ruchir said during our conversation, and a moment later, a link appeared on my phone with a detailed comparison of health insurance policies I was planning to buy.
This simple link was more than just a convenience. It transforms a scattered, opaque process into a transparent, collaborative experience for the end customer.
A distributor can also upload any existing policy document to the OneAssure AI Advisor. It could be a dense, 40-page PDF filled with legalese, and the AI can provide a concise summary. More impressively, it can answer specific, natural-language questions that the user might have about the policy.
Ruchir demonstrated this by asking the AI about a sample policy: “I have kidney stones. Can I claim on this policy?”
The AI instantly responds, explaining the concept of waiting periods for pre-existing diseases. It doesn’t give a simple “yes” or “no” but provides the necessary context for an informed decision. The distributor could ask, “What are the waiting periods on this policy?” and the AI would list them out, ready to be copied and pasted to the client.
Ruchir says, "It's humanly impossible to remember all this for hundreds of policies that an insurance agent would be servicing. The AI, on the other hand, can query 95% of the products in the market today, not just the ones we distribute, and answer any question that a buyer might have, instantly."
Crucially, this AI assistant is also embedded in the consumer-facing comparison link. This offloads a significant portion of the question-answering burden from the agent. A curious buyer can now ask the AI directly about room rent limits or specific coverages, getting instant answers and freeing up the distributor to focus on strategic advice rather than repetitive queries.
The entire journey, from initial quote to final purchase, is tracked within the platform, giving both the distributor and the customer a transparent view of the proposal status.
The Hard Part of Building AI for Insurance
"Hallucination was a big, big, big problem. We've been doing this for about eight months now. A lot of words in an Indian insurance policy are not English dictionary words, so any model at the backend will tend to hallucinate."
"Not only have we used our own proprietary data to train the models, but we've also gone ahead and used publicly available data. We've scraped everything possible to make sure that the number of times it hallucinates is the least," he says.
Engineering Speed in Insurance
One thing that stood out when Ruchir gave us a demo of the platform was its speed. When a user queries about term insurance, for instance, policies from various insurance providers appear on screen almost instantly, giving the impression of a system built for today’s impatient users. That speed wasn’t accidental. It was engineered. They overcame several external hurdles to show results quickly to the users.
One of the major technical hurdles was the unreliability of the insurance companies’ APIs, often operating in legacy systems. OneAssure built redundancy and started hosting as much information as possible on their own databases to minimize calls to external systems.
But engineering alone wasn’t enough. The team also used smart UX hacks to create a perception of speed.
"Not all insurance plans are populated instantly the moment a user clicks submit. We first show you the results of the fastest ones to load and the ones which we recommend. By the time you’re reviewing them, the rest are loading in the background," he reveals.
Beyond the Sale: Engineering Trust at the Moment of Truth
Buying insurance is only half the journey. The real test comes at claims, which is often a nightmare of delays and paperwork. This is where trust is either solidified or shattered. OneAssure has extended its first-principles approach to this critical phase.
OneAssure built a Claim Concierge to flip the traditional process where policyholders contact insurers for claims. Instead of frantic families chasing updates, OneAssure calls them, guiding them step by step with hospital paperwork, missing documents, and approvals.
"The reason why most people buy a policy is to claim eventually. What we've done is we've assured that if the claim is genuine, no matter the insurance company, we will support the customer."
"Our system knows what information is required. We let them know this is missing. That helps you fasten the process."
What’s next for OneAssure?
With over 100 active distributors on the platform and a proven model, OneAssure is now looking at its next phase of growth. The plan is twofold and ambitious.
First, they intend to unbundle their technology.
"We are going to democratize and give this platform as a Software as a Service for any distributor in the country very soon," Ruchir reveals.
This new, unbranded product will be a pure SaaS play, allowing any agent or broker in India to use their powerful AI-driven tools for a subscription fee, regardless of whether they route business through OneAssure. It’s a move that could transform them from a platform into the underlying operating system for a huge swathe of the industry.
Second, they are finally preparing to step into the D2C arena.
Armed with four years of data, refined processes, and a robust tech stack, they plan to launch a consumer-facing product within the next quarter. The goal is audacious: “If anyone wants to buy health insurance, they should consider experiencing OneAssure,” Ruchir says.
His long-term vision pushes the boundaries even further.
"I live in the future as a CEO. I have to think five years from now. Can we underwrite better? Can we look at patterns and match people to the correct products? Can I fetch your health information from your National Health ID so that I don't even have to ask you about your health conditions?"
