From IIITH to Adalat AI: building India’s justice tech stack for timely justice

Arghya Bhattacharya was 26 years old when he co-founded Adalat AI, the first of its kind LLM platform to revolutionize the Indian judicial system. Having scaled up efficiency in 20% of India’s court system, they plan to touch 50% by end 2025. Incubated by MIT, Oxford and domain leaders, it has pinged the radar of leading economists and Nobel laureates who will join the evaluation exercise.

Adalat AI in a snapshot
IIITH alumnus Arghya Bhattacharya’s Forbes 30 under 30 recognition is the latest feather in his cap. The NLP engineer co-founded Adalat AI with legal eagle Utkarsh Saxena in early 2024, to speed track justice by improving efficiency in Indian judiciary and its auxiliary infrastructure. CEO Utkarsh Saxena who featured in BW Business world 40 under 40, brings his vast experience of US and Indian legal ecosystems, and an insider’s understanding of Indian judiciary to the table.

Adalat AI is building India’s first end-to-end justice tech stack for courts. It digitizes the case lifecycle and automates manual, clerical pain points — transcription, dictation, and more..

Caseloads in Indian courts are so backed up, that one study puts the average timeline for clearance at over 300 years. Typically, High Courts’ work in English while district courts operate in regional languages.  “Our first solution was to build a legal speech-to-text Tool that could understand legal jargon, Indian accents and dialects. Our first pilot was deployed in December 2023 and our first MOU signed in April 2024”, explains Arghya, Adalat AI’s Chief Technology Officer who built the courtroom management systems. “As of September 2025, our software is in 3,000+ courts, (accounting for 15-20% of Indian courts), with official MoUs across nine High Courts.”

Early measurements indicate 30–50% speed-ups in clerical workflows like transcription and dictation in pilot courts. Broader case-resolution impacts are being studied..

Awards and finding big news on X
Arghya was at Stanford, organizing a hackathon, when he stumbled upon some heartening news! “I was scrolling X (formerly Twitter), saw a Forbes piece, opened it, and then realized my name was on the list – after first noticing Manu Chopra’s photo, since he is a friend”.  Along with making it to the Forbes List, Adalat AI has received international recognition at prominent events like Fast Company’s ‘Next Big Things in Tech 2024’, ‘4 Best International Tech Innovations of 2024’ and MIT Solve.

LTRC – where it all began
“I actually owe a lot of what I am doing now to IIIT Hyderabad”, says Arghya who comes from a background where his father was in a transferable role, building metros and bridges across the country. With school education spanning India, he speaks many languages including English, Bengali, Hindi, Oriya, Tamil and a smattering of French and Telugu.  

Arghya first learnt about IIIT Hyderabad’s dual degree CLD program while preparing for JEE. “Given my background in languages, my education at IIITH almost feels God-sent. I still remember Orientation day when Prof. Dipti Misra explained their vision for LTRC, and how computers that interpret language will soon unlock a whole different paradigm of computing. Back in 2016 no one had an idea about ChatGPT! CLD literally put us at the center of the current AI boom. In many ways, the credit for where I am today goes to the makers of the course, and professors like Prof. Radhika Mamidi and advisor Prof. Manish Srivastava who spurred me on,” observes Arghya, first in his batch to publish in A* conferences. His current work is an extension of his academic research on advancing aspect-based sentiment analysis for low resource languages, and developing a novel architecture in a multilingual setting.

A wide spectrum of activities made college life very vibrant for the young engineer. Google Summer of Code, an award for athletics and football were a big part of his IIITH memories.  “Though I was an active member of the placement committee, I never applied for jobs myself since entrepreneurship was always my endgame”, notes Arghya who pegs the rejuvenating breakfasts at Yuktahar Mess, Felicity and hanging out with seniors as treasured moments. “My closest circle is my IIIT group of friends, whether it is for holiday trips or work visits in India and abroad”.   

Breaking Language Barriers after IIIT Hyderabad
After productive internships at ISB and Sentisum, Arghya began his professional journey at Enterpret, one of the most reputed startups to come out of IIITH. Three years later, he moved to Hyderabad-based Equal, to work on building Indian facial recognition models. An overarching desire to work on Bharat/India-specific impact-first projects saw the engineer quit his job on New Year’s eve 2023.

Earlier in 2023, Utkarsh Saxena, a Supreme Court advocate was hit with the Adalat AI brainwave. Things started moving fast when Arghya responded to a WhatsApp message calling for AI experts.  When Utkarsh sketched out the reasons for the glacial pace of Indian courts and the scope for tech innovation in judiciary, Arghya realized that he had finally landed on his dream project. That was the Aha moment for Arghya who believed that in the age of Aadhaar and UPI, Courts should not be left behind.

How the start-up grew on goodwill and funding
“I still remember my first visit to a court; paper-laden records rooms with clerks manually locating files”, recalls the entrepreneur. “In many courts, there is a widespread shortage of stenographers that is forcing judges to manually take notes. With Adalat AI, we are the first in the world to enhance productivity in courtrooms using software”.

The clincher was having a supreme court advocate who understood the intricate workings of the Indian legal system.  “Given the sensitivity of court data, we build and deploy our own models on in-country infrastructure. Data remains encrypted and is accessible only to the Courts; no third parties have access.”.  We took our Figma files to Courts, received constant feedback from Justices and tweaked the architecture. Legal experts were co-creating the model with us and championing its implementation as well”.

Incubated at Harvard, Oxford and MIT, Adalat AI’s journey was supported by Microsoft’s Founders Hub, under the IndiaAI mission. “Our advisory board has a mixed bag of eminent retired Chief Justices, High Court judges of Bombay and Orissa, associate professors from MIT, UPenn and Stanford, an ex-CEO of Infosys, and the HOD of Stanford CS”, reports Arghya. Non-profit tech accelerators – Fast Forward and Nudge were among the first evangelists who speed tracked their journey along with leaders Iike Aziz Gupta, CEO Rocket Learning and IIITH alumni Vamshi Ambati. The fully remote operation with a 30 strong team, comprises of top talent from India and around the globe, who left high-paying jobs to move back to India to work in this sensitive domain, in the non-profit space. 

“Our evaluation approach has drawn interest and input from leading economists (including Nobel laureates). We will share details once formalized.””, observes Arghya. Judicial systems in Ghana, Nigeria, Kenya and Bangladesh have expressed interest in the technology. “Most colonized nations of the Global South have similar legal structure, jargon and problems and thus solutions that we roll out can be standardized.  In fact, it was quite surreal to see how closely the court in Ghana resembled the Bombay High Court!”

Sports mentality brings out the best
“Two great influences in childhood was my parents’ loving and respectful relationship and the keeda for doing something in the social space, that came from my mother”, remarks Arghya, who was a keen sportsman, state level athlete and captain of his football team in school. “My attitude to sports has shaped who I am as an adult and defines my work ethic. On a 10 kilometer run, when your brain is telling you to give up at that 7th kilometer, it is your self- talk that will help you finish the race. It is this grit that translates into the workspace. Though we are a non-profit, we work like a professional sports team”, quips the data nerd who tracks everything in his life, from sleep, food, fitness and emotion spikes on his Whoop band.

“I see failure as feedback”, philosophizes Arghya who learnt the art of patience from Utkarsh and meditation from his mother, “who is pursuing her Ph.D in cognitive behavioural therapy”. He feels most spiritually connected, when mountain climbing, scuba diving, on the motor sports track or cross-country running. Today, sleeping in his own bed and spending time with family and Leo his golden retriever is what sparks joy.  Jazz tops his current music playlist along with books “that fire the neurons of my brain in ways which my job does not allow me to do, picking up mental models from domains, like history or psychology” says the occasional percussionist who loves his drum kit.  

Among works in the pipeline is a mobile-based digitization software for handwritten documents and a visit to the UN General Assembly.  “We are trying to do to the legal ecosystem what UPI did for payments — make the basics fast, reliable, and universal. Court data is sacred and that is why we build in-country, privacy-first systems with no third-party access. If we’re lucky enough to dent this, I’ll move on to the next hard problem”, smiles Arghya Bhattacharya.