As India's digital landscape expands beyond metro cities, regional-language social platforms are redefining how millions connect online. In this exclusive conversation with Rajeev Ranjan, Editor, Digital Terminal, Harshvardhan Chhangani, Co-Founder and CPO, FRND (Interact Group), discusses building a voice-first social platform for Bharat, leveraging AI for safer and more meaningful interactions, tackling large-scale multilingual challenges, and balancing technological innovation with authentic human connections.
Rajeev: FRND is building for users beyond metro cities, where digital behavior and social needs are often very different. What product and technology decisions have been most critical in creating an experience that resonates with Bharat users?
Harshvardhan: Building for Bharat users meant starting from a fundamentally different set of assumptions. For our community, trust is enabled through every interaction with the product. So our priority was making the experience genuinely seamless, which included intuitive navigation, easy-to-comprehend features, and no moments where a user feels lost or unsure. We also built in product nudges and safety guardrails as core design elements that helped users engage in the right way and feel assured that the space they were entering was safe and authentic.
On the product side, two decisions stand out. First, we made voice the primary mode of communication. Our users aren't always comfortable with text. Voice is more natural, familiar, and expressive in a way that typing simply isn't for many people in Tier 2 and Tier 3 India. Second, we were very strict with moderation from the very beginning. We set clear rules around what behaviour was acceptable on the platform and acted swiftly against users who broke them.
Every product decision, from feature design to discovery flows to the overall user journey, has been made with this user as the north star.
Rajeev: Unlike traditional social platforms that are video-centric, FRND has adopted an avatar-based, audio-first approach. What was the thinking behind this model, and how has it impacted user adoption, privacy, and engagement?
Harshvardhan: We noticed early on that a lot of users from Tier 2 cities and beyond were hesitant to put their real photos on the platform. There's a social conservatism that exists in these communities, and for women especially, there's still a real stigma attached to being visibly present online. An avatar-based identity removes that barrier entirely. It meant that our community could show up, express themselves, and connect without any inhibitions.
Rajeev: Trust and safety are fundamental to social discovery platforms. How has FRND architected its moderation and safety framework to create a secure environment, particularly for first-time internet users and women?
Harshvardhan: FRND is a synchronous platform, this means that there are live interactions taking place. Hence the need for robust AI systems and real time moderation becomes critical to maintain user confidence.
Our moderation process follows a four step approach.
Human-led moderation
We have a dedicated team of 100+ human moderators who actively monitor the platform. They conduct behavioral compliance inspections in rooms and visit rooms to observe user behaviour, ensuring compliance with our community guidelines & terms of use and taking appropriate action when necessary.
Safety-by-design features
Our safety-by-design approach means every surface is engineered to be intuitive, inclusive, and safe. We have a lot of in-app features that prompt users to be respectful and keep the platform free from bad behaviour. These include a lot of thoughtful nudges and design features. Listing a few of them below
Users are required to verify their gender by reading out the text written on the screen which is verified by a gender attributing AI. Recorded voices or fake profiles are not allowed.
They are then allotted a custom avatar which becomes their identity on the app which enables privacy and also keeps users safe from misuse of their personal pictures.
Every user who redeems their virtual gifts to cash after accumulating INR 100 in total redemptions and withdraws it to their UPI wallet has to go through FRND’s KYC (Know Your Customer) process. The users who fail to pass the KYC checks & are listed as risky accounts by our internal audit systems are not allowed to withdraw money to their wallets.
Once inside the rooms, we display safety education cards which set the tone for a safe and dignified conversation. The user is warned to avoid engaging in sleazy conversations via an interactive pop-up. RJs/Chaperons are warned that partaking in rude or abusive conversations might lead to lower ratings which might deteriorate their experience on the app.
In video format rooms it is impossible for any violator to perform sexually explicit displays during the video conversation. This is ensured by an AI face detection algorithm. It is mandatory for all users to display their faces while having a conversation. Failure to show face results in automatic & instant blurring of the video stream effectively turning the call into only audio until both the host & the user position their camera such that their faces become visible again.
To counter extortion, FRND continuously educates users via informative pop-ups about not sending money to any other user off the app & to not share their personal details.
Taking screenshots of video calls in the app is prohibited to protect the privacy of users.
Users who have been found to violate Platform Policies (Community Guidelines, Terms of Use) are given strict warnings by FRND Moderators. In a case where the user warned by the FRND Moderators repeats the violation, they are temporarily or permanently blocked from using the app, based on the severity & frequency of their violations.
Reactive moderation
Users can kick out/report misbehaving users anytime via upfront reporting features. Reports are reviewed by our human moderation team for appropriate action. If a RJ or user prefers not to connect with someone, they won’t be matched again, ensuring a more personalised experience. The moderation team also takes reports into account when they warn or temporarily/permanently block the users. The reporting option is prominently visible in all calls.
For us to maintain the sanctity of the platform while our community indulges in interactions, it is absolutely essential to weed out the bad actors. Prioritising our trust and safety vertical gives us a competitive moat.
Rajeev: AI is increasingly shaping digital experiences. How is FRND leveraging AI across user discovery, content moderation, matchmaking, and engagement while ensuring authenticity in human interactions?
Harshvardhan: We have always seen AI as an enabler and to lower the friction for human connection and not replace it, we believe that as AI becomes more abundant, humans will crave for interpersonal connections and we are solving for that innate human need.
On the moderation side, we adopted AI-driven systems early even when we were a relatively young company. That was a deliberate call. Building this capability in-house at an early stage gave us a meaningful technical edge and pushed our team to innovate faster. AI also helped us to act in real time flagging harmful behaviour before it escalated.
On discovery, our AI matches users across the same Indic language, helps them discover relevant communities, connection. It reads conversation signals, understanding how users interact making matchmaking between two users more natural and authentic.
Rajeev: With millions of users interacting across multiple Indic languages, what are the key technology challenges in delivering seamless matching, room discovery, and session continuity at scale?
Harshvardhan: One of the unique challenges of building a social platform for India is that you're operating across multiple languages, cultures, and network conditions simultaneously. Delivering a great experience isn't just about matching users with relevant people and rooms; it's about doing so reliably in real time for millions of users.
On the discovery side, the challenge is balancing relevance and availability. We need to help users find conversations that match their language, preferences, and intent while ensuring the ecosystem remains dynamic and fair for both new and established creators.
At the infrastructure level, session continuity is critical. Mobile connectivity in India can be highly variable, so the platform needs to gracefully handle network fluctuations and transient disconnects without disrupting the user experience. For users, conversations should feel seamless even when underlying network conditions are not.
Another challenge is operating at high concurrency. During peak periods, large numbers of users may be trying to join the same experiences or discover new conversations at the same time. Ensuring reliability, fairness, and low latency under these conditions requires significant investment in real-time systems and distributed infrastructure.
Ultimately, our focus is on making the complexity invisible to the user. Whether someone is joining a room in Hindi, Tamil, Telugu, or any other language, the experience should feel instantaneous, relevant, and dependable regardless of where they are or the quality of their connection.
Rajeev: As AI-driven interactions become more common, how do you see the future of social platforms evolving? Will technology strengthen genuine human connections, or risk replacing them, and how is FRND positioning itself for that future?
Harshvardhan: We don't see AI and human connection as opposing forces. In fact, we believe AI will become a powerful enabler of social experiences. AI can help users discover the right people, find relevant communities, overcome language barriers, advance safety mechanisms, and make digital interactions more personalized and engaging.
As AI becomes more deeply embedded into everyday experiences, authentic human relationships will become more valuable. People will continue to crave belongingness, shared experiences, cultural context, and emotional connection.
That's where FRND is uniquely positioned. We bring together people who share a language, a cultural context, and common interests. Whether it's discussing a favourite movie, celebrating a festival, participating in a group conversation, or simply finding audio companionship, technology helps facilitate the connection, but the value ultimately comes from the humans on the other side.
Our approach is simple: we are excited to incorporate AI wherever it can meaningfully improve the user experience, but we remain focused on the underlying human need of ‘interaction’ that we're serving. At its core, FRND is a positive digital third space where users can express themselves, build relationships, and find a sense of belonging, with AI helping make those experiences more accessible, relevant, and safe.
𝐒𝐭𝐚𝐲 𝐢𝐧𝐟𝐨𝐫𝐦𝐞𝐝 𝐰𝐢𝐭𝐡 𝐨𝐮𝐫 𝐥𝐚𝐭𝐞𝐬𝐭 𝐮𝐩𝐝𝐚𝐭𝐞𝐬 𝐛𝐲 𝐣𝐨𝐢𝐧𝐢𝐧𝐠 𝐭𝐡𝐞 WhatsApp Channel now! 👈📲
𝑭𝒐𝒍𝒍𝒐𝒘 𝑶𝒖𝒓 𝑺𝒐𝒄𝒊𝒂𝒍 𝑴𝒆𝒅𝒊𝒂 𝑷𝒂𝒈𝒆𝐬 👉 Facebook, LinkedIn, Twitter, Instagram
