Pakistani Legal AI Platform Creator Talks Details About Qazi AI

Qazi.ai is an upcoming AI platform built to provide legal help and guidelines to casual and professional users alike. The platform is currently in its alpha testing phase, but we have reached out to its creator Aadil Jaleel for an interview that answers in-depth questions about Qazi AI and how it could revolutionize Pakistna’s judicial system. Qazi.ai is an AI-powered legal technology platform designed to streamline and transform Pakistan’s judicial system. The project was founded by developers Aadil Jaleel and Daniyal Yasin, both of whom are currently heading vResolv.IO. With over 2.2 million pending cases across the country and 57,000 cases stuck in the Supreme Court alone, Pakistan’s legal system struggles with inefficiencies, delays, and a lack of automation. Qazi.ai addresses these issues by revolutionizing the entire judicial process—from case filing to legal research, mooting, document drafting, and even judgment generation. The platform integrates AI to assist legal professionals, courts, and the general public in navigating the complex legal landscape with greater ease and accuracy. The idea for Qazi.ai was born out of first-hand frustration with many problems and inefficiencies in the legal system. As for the name “Qazi”, it is inspired by the Ottoman Qazis, known for their just and efficient judiciary. Qazi.ai is designed to accelerate justice, increase court accessibility and credibility, and democratize legal knowledge. Thus, it operates on a B2B, B2G, and B2C model, ensuring that its impact reaches legal professionals, government institutions, and the general public. For lawyers and law firms (B2B model), Qazi.ai provides an end-to-end case management system that streamlines legal research, mooting, precedence analysis, and automated legal document generation. For government institutions (B2G model), Qazi.ai helps increase court accessibility and credibility by streamlining case processing, judgment drafting, and legal automation. For the general public (B2C model), Qazi.ai works to democratize legal knowledge by making legal information accessible and understandable. Qazi.ai bridges provide simplified legal language, preliminary legal guidance, and basic legal document drafting, making legal resources more available to everyone. Developing Qazi.ai in Pakistan came with significant challenges, largely due to structural inefficiencies in the legal sector. One of the biggest obstacles was data availability and digitization. Unlike many developed countries, Pakistan lacks centralized, structured, and digitized legal databases. Legal records are often fragmented, handwritten, or difficult to access, making it challenging to train AI models on local case law. From a technical standpoint, AI infrastructure and costs posed hurdles Legal AI requires high-performance computing (HPC), GPUs, and cloud storage, all of which come with significant costs (more on that later). Running large-scale AI models is expensive, and local AI infrastructure remains limited, increasing reliance on external cloud solutions. Beyond technical challenges, there is also a general resistance to change. The legal profession is deeply rooted in traditional methods, and many professionals are hesitant to trust AI-driven legal assistance. Additionally, intellectual property (IP) theft is a real concern. The cost of developing and running an AI-driven legal platform like Qazi.ai is significant, given the complexity of the system and the scale of stakeholders it serves. The development, deployment, and launch of Qazi.ai for all three key stakeholders—judiciary, law firms, and the public—can cost up to Rs. 220 million ($800,000). This includes data collection and preprocessing, AI model training, system integration, and UI/UX development. The running costs primarily involve infrastructure expenses, especially when deploying on cloud platforms. These costs depend on the scale of adoption and deployment strategy. If hosted on cloud services, monthly costs can range from Rs. 22 million to 65 million, scaling from 10,000 to 100,000 users. Qazi.ai is built on a Retrieval-Augmented Generation (RAG) architecture, leveraging  Meta’s LLAMA open-source model as its foundational large language model (LLM). Open-source models provided significant advantages, particularly in terms of data security, privacy, and customization. While we remain committed to open-source solutions due to their transparency, flexibility, and cost-effectiveness, we are always evaluating trade-offs. If a major shift occurs in performance, cost, or efficiency, we will reassess our approach to ensure we are using the best possible technology. Qazi.ai is currently in its alpha stage, where it is being tested and refined through invite-only access. The focus is on ensuring that the system meets the practical needs of legal professionals and institutions. We are working towards completing and deploying a market-ready product soon. To accelerate development and scale deployment, we are also in the process of raising funds. Qazi.ai is not just a legal automation tool—it is part of a broader vision to modernize and digitize Pakistan’s legal ecosystem. In the near future, we plan to expand Qazi.ai’s capabilities by introducing: In the long term, we envision Qazi.ai becoming a regional leader in legal AI, expanding into the US market in the future. To aid these ambitions, QaziAI is also pitching its product at the Meet The Drapers competition for legal startups created by Silicon Valley venture capitalist and entrepreneur Tim Draper. The competition is based on public votes and you can cast yours here. Simply scroll to the “AI Tiktok” section and select Qazi AI.