April 2026 Summaries

In 2026, telehealth platforms evaluating video infrastructure options must balance compliance readiness, developer velocity, and cost predictability, with VideoSDK emerging as a strong contender for startups and growth-stage platforms due to its ease of integration and managed low-latency infrastructure. Meanwhile, Zoom SDK appeals to enterprises with existing Zoom agreements due to its brand familiarity and HIPAA compliance, though it offers limited customization and incurs potential vendor lock-in. Raw WebRTC provides maximum control over the media stack, making it suitable for large platforms with engineering expertise and strict data residency requirements, but it demands significant development effort and assumes all compliance responsibilities. The decision between these technologies hinges on a trade-off between infrastructure control and time to market, with each option evaluated based on integration complexity, compliance, latency, scalability, cost, and vendor dependency.

Patient drop-off during telemedicine video calls presents a significant challenge, often resulting from poor onboarding, device readiness issues, and waiting room abandonment. To address this, telehealth platforms should conduct a comprehensive audit of the entire patient visit funnel, from appointment booking to post-call completion, and implement targeted optimizations at the highest-friction points. Key interventions include pre-call device checks, simplified authentication, and robust video infrastructure to prevent connection failures. Most platforms face a 20–40% drop-off rate, which can be reduced through structured onboarding processes, real-time status updates in waiting rooms, and browser-based access that eliminates the need for app installations. By focusing on each stage of the funnel independently, platforms can enhance patient retention and improve both revenue and health outcomes in an increasingly competitive telehealth market.

The Insurance Regulatory and Development Authority of India (IRDAI) mandates the Video Based Identification Process (VBIP) for remote insurance customer onboarding, requiring live video verification to ensure customer identity before issuing policies digitally. This process necessitates InsurTech companies and insurers to develop compliant real-time video infrastructure that includes capturing geo-tagged consent, recording sessions, and maintaining auditable storage in line with IRDAI guidelines. VBIP replaces traditional in-person KYC by verifying identity through live video interactions while ensuring document authenticity and capturing informed consent. This framework applies across all insurance product lines and is integral for fraud prevention in digital insurance distribution. Compliance entails adherence to specific technical and procedural requirements, including session recording, geo-tagging, and data retention, with significant penalties for non-compliance, ranging from regulatory fines to potential licensing risks for insurers and associated InsurTech platforms. The process also intersects with other regulatory frameworks like Anti-Money Laundering (AML) and Aadhaar-based verification, requiring meticulous coordination to ensure adherence to all relevant regulations.

Scaling video-based KYC systems to handle over one million verifications monthly involves intricate infrastructure planning, particularly for financial entities in India due to regulatory requirements like RBI's V-CIP. The architecture necessitates a robust distributed WebRTC setup capable of managing 1,500–2,000 peak concurrent sessions and an agent model supporting 8-10 verifications each per hour. Compliance with RBI mandates includes live, uninterrupted video interactions, geo-tagging, and secure storage within India, with penalties for non-compliance. Globally, similar digital identity verification standards are emerging, further complicating infrastructure needs. Key challenges include managing peak concurrency, ensuring recording integrity, and maintaining adequate trained agent capacity. Hybrid architectures often provide a balanced approach by using real-time video infrastructure SDKs for session management while retaining control over the KYC backend and compliance components. Effective failure management, such as handling network drops and ensuring recording redundancy, is essential for maintaining audit trails and compliance.

Video KYC (Video-based Customer Identification Process) has become a crucial method for onboarding in India's financial sector, following guidelines set by the Reserve Bank of India. Despite its regulatory importance, the process is plagued by high abandonment rates, with typical completion rates between 55-75%. The main drop-off points occur during pre-session readiness, document capture, and queue wait stages. Addressing these through improvements in user experience, infrastructure reliability, and compliance with RBI mandates can elevate completion rates to over 80%. Key strategies include offering pre-session guidance, enhancing document capture with real-time feedback, providing queue transparency, and ensuring session recovery. Successful optimization requires a systematic approach that considers the technical, procedural, and user-related aspects of the process, with a focus on ensuring compliance without compromising user experience.

In 2026, regulated fintech companies and non-banking financial companies (NBFCs) in India must navigate a complex landscape of Know Your Customer (KYC) methods, each with distinct compliance, scalability, and fraud resistance profiles. Video KYC, conducted via live, agent-assisted video calls, offers robust compliance and auditability, making it the preferred choice for high-value financial products, particularly where Aadhaar-based eKYC is unavailable or consent cannot be obtained. eKYC, leveraging Aadhaar data for identity verification, is the fastest and most scalable method but is limited by regulatory constraints and the necessity for customer consent. Physical KYC remains relevant only in instances where digital infrastructure is inadequate or specifically mandated by regulators. The Reserve Bank of India's guidelines provide a framework for these methods, emphasizing secure data handling, compliance requirements, and the potential penalties for non-compliance. The choice of KYC method impacts operational costs, customer acquisition speed, and fraud risk, with video KYC being favored for its balance of digital reach and regulatory adherence.

SEBI-regulated stockbrokers in India are required to complete an In-Person Verification (IPV) for every new account, with video-based IPV emerging as the primary method, though distinct from the banking sector's "Video KYC." This compliance process involves a live video session to verify the investor's identity against existing KYC records, such as PAN, Aadhaar, and CKYC/KRA, and mandates storing a tamper-proof recording for audit purposes. Governed by SEBI's KYC norms and PMLA obligations, the process is separate from the RBI's framework applicable to banks. The guide emphasizes the importance of distinguishing between SEBI's video-based IPV and RBI's Video KYC, providing a detailed overview of the end-to-end onboarding process, including authentication, CKYC lookup, video session compliance, and the critical nature of recording infrastructure. It also discusses the build-versus-buy considerations for implementing video IPV systems, recommending a hybrid approach for mid-size brokers to balance compliance and infrastructure scaling needs. Compliance officers, CTOs, and product leads must ensure adherence to SEBI's detailed requirements to avoid significant penalties and reputational damage.

Prism: Agents V1.0.0 represents a comprehensive overhaul of the VideoSDK AI voice pipeline, addressing limitations found in previous versions. The update introduces a unified Pipeline class that replaces the previous CascadingPipeline and RealtimePipeline, allowing for greater flexibility and interoperability between components such as speech-to-text (STT), language models (LLM), and text-to-speech (TTS) systems. This new architecture supports three modes—Cascade, Realtime, and Hybrid—enabling customization and integration with various models and tools, while also enhancing observability with built-in metrics, logging, and tracing. The system is designed to run concurrently, reducing latency and improving responsiveness, and includes a robust hooks system for data interception and transformation at different stages. The release includes 22 production-ready templates for common use cases and maintains compatibility with existing tools and lifecycle management features, providing a scalable and flexible framework for AI voice agents.

In the March edition of VideoSDK Monthly Updates, the company announces the launch of Agents v1.0.0, marking a significant milestone for their AI stack with a complete architectural shift in building AI voice agents. This new version introduces a unified pipeline architecture that replaces previous models with a flexible system that adapts automatically to various use cases, including Cascade, Realtime, and Hybrid modes, all of which enhance the customization and integration of AI components. The update also introduces a hooks system for data transformation, observability features for monitoring, and new plugins like Anam AI Avatar and LangChain support. AI agents now have native support across all VideoSDK platforms, allowing for real-time transcription and state tracking, alongside the integration of Google Gemini 3.1 for ultra-low-latency responses. The update is complemented by the release of starter apps and new guides to assist developers in deploying and integrating these advanced AI voice agents.