個案總覽

Jugalbandi is an AI-driven platform developed by OpenNyAI, aimed at enhancing access to legal information and government services in India. This innovative platform integrates large language models (LLMs) like ChatGPT with language translation models (LTMs) from Bhashini, creating a seamless conversational AI interface. OpenNyAI, initiated by Agami, leverages AI to address societal challenges, particularly within the legal domain. Jugalbandi was conceived to mitigate the extensive backlog of pending legal cases in India by making legal information more accessible and comprehensible to all, especially in local languages. The platform has gained significant recognition, including being highlighted by Satya Nadella during Microsoft Build 2023, and has been instrumental in delivering critical information through popular messaging apps. Despite its success, Jugalbandi faces several strategic challenges that need to be addressed to ensure its scalability and sustainability. One major dilemma is balancing ecosystem development with technological dependencies on external AI models like ChatGPT and Bhashini. Ensuring the platform remains up-to-date and effective while relying on these foundational technologies is critical. Another significant dilemma is choosing between maintaining an open-source model, which aligns with the mission of creating digital public goods (DPGs), and transitioning to a for-profit model to ensure financial sustainability and growth. Additionally, the team must decide between vertical diversification, adding more core features to enhance existing applications, and horizontal diversification, expanding to new use cases across different domains. Each approach has implications for resource allocation, community engagement, and technological development. Addressing these dilemmas is crucial for the long-term success and impact of Jugalbandi.

The PM begins with exploring the problem space, gaining an understanding of the larger context of the problem, namely the online bus travel sector in India and the challenges faced by the industry players. Then he conducts a deeper investigation of the actual user behavior data on redBus using Product Analytics (specifically, the method of funnel analysis) to identify the major scenarios when the platform users were not completing the ticket purchase. The PM uncovers that the users searching for bus tickets on the shorter routes seemed to drop off significantly more just after looking at the available bus options, as compared to the longer routes. Next, the PM formulates hypotheses to examine the reasons behind this intriguing behavior of short route travelers, and tests them using customers interactions and field studies. From the rigorous user research and hypotheses testing, the PM could confirm three categories of expectations of short route travelers which were not being fully met by the current features of the platform. Armed with this clarity about the problem space, the PM moves into the solution space, ideating feature solutions for the users. With inputs from various stakeholders, he finally designs a platform feature called 'Open Ticket' that could fulfill the three needs of their short route users. Then he worked with the engineering team to develop and launch the MVP of the solution using the Agile Scrum methodology. Finally, the post-launch user behavior was tracked to validate the efficacy of the Open Ticket solution. Unfortunately, the designed solution did not work as per expectations, leaving the PM with a decision dilemma that is critical in lean startups, i.e., whether to persevere with the Open Ticket idea by trying to fix it, or drop it completely and pivot to a new idea?