Kata.ai, one of Indonesia’s most prominent artificial intelligence (AI) innovators, has carved a unique niche in conversational AI by offering solutions that blend advanced natural language processing (NLP), voice recognition and AI-driven automation. CEO Irzan Raditya shared how Kata.ai has transformed from a consumer-focused personal assistant into a sophisticated business-to-business (B2B) AI platform used by financial institutions, telecommunications providers and governments across emerging markets. “We started as a personal concierge service over short message service (SMS),” Raditya said. “People could book flights or restaurant tables,and we managed everything manually, with about 50 human agents.” But the team saw an opportunity in leveraging their growing dataset to build machine learning models for natural language processing, long before the AI boom driven by tools like ChatGPT. Expanding the future of conversational AI in financial services In 2017, Kata.ai made a decisive pivot to become an AI company focused on enterprise clients. “We began building conversational AI that could understand and process Bahasa Indonesia,” Raditya explained. “That was our starting point before expanding to other languages.” Today, Kata.ai supports over 100 enterprise clients, with a significant portion in financial services. Its solutions range from chatbots to voice assistants and digital avatars, all designed to enhance customer experience and operational efficiency. “We are now top of mind for conversational AI solutions in Indonesia,” Raditya affirmed. From chat to voice and digital avatars to meet the evolving needs of banks Kata.ai’s initial success with chatbots on platforms like WhatsApp led to the expansion into voice services, an area that has seen rapid uptake among financial institutions. “Voice is increasingly popular, especially in collections and customer service,” Raditya noted. For example, Kata.ai recently deployed a voice AI system for PT Kereta Api Indonesia (KAI), Indonesia’s state-owned railway company, to handle customer inquiries and service requests through voice interactions. However, the company’s most groundbreaking offering is its digital avatar technology, already being piloted by banks and service providers. One notable implementation is with Pegadaian, Indonesia’s government-owned pawnshop and financial services provider. Pegadaian uses Kata.ai’s digital avatars in branches to deliver personalised customer service on gold investment products and loans. “These avatars act as virtual customer service representatives, available 24 by 7, capable of conversing in multiple languages and providing a lifelike customer experience,” Raditya said. During the demonstration, the avatars convincingly handled complex customer queries, offering an experience that blurred the line between human interaction and AI assistance. Kata.ai is also advancing its agentic AI capabilities—AI agents that proactively interact with customers and make decisions based on context. “A customer can upload a picture of a car they want to buy, and the AI agent identifies it, estimates its price, and builds a personalised savings plan based on the customer’s goals,” Raditya explained. This multimodal approach, combining image recognition and natural language processing, creates a seamless and intuitive customer journey. Navigating AI regulation and data privacy in financial services As AI becomes deeply embedded in banking operations, issues of data privacy and regulatory compliance have taken centre stage. Indonesia, like many countries, mandates that financial institutions store and process data locally. This has led to growing demand for private, on-premise AI deployments that satisfy data sovereignty laws. “Banks need private GPT solutions deployed on-premise,” Raditya emphasised. “We help financial institutions fine-tune large language models (LLMs) on their own datasets while ensuring full compliance with data regulations.” Kata.ai uses a hybrid approach, combining proprietary and open-source language models depending on the client’s needs. It has worked with OpenAI and Google for certain projects but prioritises open-source models like Meta’s Llama for deployments requiring tight control and local processing. Although Kata.ai has explored Chinese-developed AI models, such as DeepSeek, Raditya acknowledged concerns about latency and performance compared to proprietary offerings. The company also offers private GPT solutions for financial institutions, enabling them to create AI systems that draw insights from internal datasets while maintaining strict data governance. “This allows banks to query complex data in natural language securely,” Raditya said, adding that Kata.ai’s systems are designed to integrate easily into legacy information technology (IT) infrastructures. Enabling omnichannel AI deployment across platforms and regions Kata.ai’s conversational AI solutions are designed for omnichannel deployment, ensuring a seamless experience whether customers interact via mobile apps, websites, call centres, or physical branches. “We’re providing AI services that banks can deploy in-branch through kiosks and digital screens or online through their apps and websites,” Raditya explained. This flexibility has been key to Kata.ai’s expansion beyond Indonesia. The platform now supports more than 30 languages, opening up markets across Asia and beyond. In addition to financial services, Kata.ai is also expanding into telecommunications and public sector deployments. Its scalable architecture enables institutions of all sizes—from multinational banks to regional fintechs and state-owned enterprises—to deliver consistent and personalised customer experiences. A profitable and self-sustaining business with ambitious plans Kata.ai’s ability to attract significant enterprise clients has translated into a profitable and sustainable business model. Early investments came from Trans-Pacific Technology Fund, a Taiwanese tech investment firm that led Kata.ai’s Series A and Series B funding rounds. Other backers include AC Ventures, MDI Ventures (Telkom Indonesia’s VC arm), and investors from Japan and South Korea. “We’ve been profitable for a while now, and most of our revenue comes from customers,” Raditya confirmed. Financial services clients make up around half of Kata.ai’s customer base, with the remainder coming from telcos, retail and government sectors. As part of its growth strategy, Kata.ai has raised $5 million so far and plans to do more to expand its market reach and invest in research and development. The company aims to further enhance its agentic AI capabilities and develop new multimodal solutions that incorporate voice, text and image recognition technologies. Kata.ai also plans to expand partnerships with cloud providers and system integrators to make its AI solutions more accessible to financial institutions in emerging markets. “We’re committed to enabling banks and businesses in Asia and beyond to tap into the transformative power of AI,” Raditya said. Helping financial institutions unlock AI’s potential Kata.ai’s evolution from a personal assistant app into a cutting-edge conversational AI platform illustrates how strategic innovation and customer-centric design can drive digital transformation in financial services. Its blend of multilingual chatbots, voice AI, digital avatars and agentic AI systems is helping banks and other institutions deliver more personalised, efficient and scalable customer experiences. “Our goal is to help banks and financial institutions kickstart their AI transformation,” Raditya concluded. “Whether through private GPT deployments, AI-powered customer service, or digital avatars, we enable our clients to deliver smarter and more personalised experiences to their customers.” With a focus on regulatory compliance, regional expansion and continuous innovation, Kata.ai aims to play a pivotal role in shaping the future of conversational AI in emerging markets.
