In a packed chamber in Delhi, a senior counsel leans back in his chair, eyes closed, dictating a petition draft to his junior. Names, citations, procedural steps—all flow with the authority of experience. But across the desk, the junior is perplexed. “Did I hear Section 482? or was it 432?”, he reflects. A moment passes, he hurries through the draft. Something gets typed in. All looks good – for now.
In another courtroom, a judge reads out a detailed order. The stenographer listens intently, typing furiously, tries to capture every word in shorthand— a misheard phrase or a skipped citation—”persecution” becomes “prosecution” and distorts fundamental meaning. Weeks later, a litigant is denied relief not on merits, but because the order’s wording is imprecise. No one notices the missing phrase until a litigant’s liberty is lost to a technicality.
This is a real problem. In 2021, a similar incident happened in Gujarat. The accused was incarcerated despite being granted bail by the Gujarat High Court simply because of a typo – incorrectly listed petitioner numbers in the court’s written order. The intent was clear, but the mistake meant that prison officials refused to release the accused until a formal correction was issued. You can find the original order referenced here and the news report regarding the typographical error here.
This is the hidden crisis of Indian legal practice—where dictation and transcription costs access to justice. Junior advocates struggle with context, stenographers miss nuance, and generic voice tools fall apart when faced with legal complexity in an India-sized problem.
Indian legal practice presents a curious paradox: while research and case management have embraced digital transformation, dictation – the backbone of legal drafting - remains frustratingly analog. Even as e-filing becomes the norm and courts go virtual, most advocates still struggle with unreliable recording devices or generic speech-to-text tools that can’t quite keep pace with the Indian legal lexicon.
jhana steno, a new entrant from legal tech startup jhana.ai, promises to end this frustration. India’s first dictation tool, built for the law, it quietly addresses the real-world pain points that every judge, litigator, corporate lawyer, or chamber junior has felt for years.
Generic voice software and legal practice mix about as well as oil and water. Ask any lawyer who's tried dictating 'Writ of Mandamus' and got 'Rit of Man Thomas' in return. jhana steno is different. From the first use, it becomes clear that this tool “thinks” in legalese. Dictate a statute or citation, and it’s automatically formatted to the standards of any major court. Case names are italicized; paragraph numbers fall neatly into place; and grammar is polished on the fly.
New multimodal LLMs are already far more capable at recording dictation in a manner that captures the full intent, style, and meaning of the spoken draft. But this is just the beginning–a real stenographer also edits, auto-completes citations, makes corrections, produces legal formatting, and yields a truly interactive experience. You can ask for anything from a year being filled in a citation to a formatting change a few paragraphs before. So, steno leverages the voice AI advances in ChatGPT, Gemini, etc. but takes them further. It adds multiple transcribers backchecking each other’s work, legal editors fixing typesetting and checking legal grammar, and revision agents that fix any issues. And at any point, you can pause your transcript and manually edit the draft. Or you can export it to MS Word, preserving all the formatting and intelligence. And whenever you please, resume your drafts right on steno.jhana.ai, leveraging the ability of steno to revise or change prior text just as it can draft new text! This means steno is more than a legal version of voice AI systems–it is a series of latest AIs collaborating to ensure the best possible output with the least possible input, and it is a complete MS Word like interface where humans and AIs drafttogether.
“Frankly, I expected yet another gadget that would make me do extra work,” admits a senior Delhi-based counsel Abhisek Kumar. “But (jhana) steno actually gets the nuances. I spend less time fixing errors, and more time focusing on strategy.”
The magic lies in its Indian DNA. The team behind jhana steno trained the tool not just on Indian voices, accents, regional inflections, and courtroom echoes but also on real legal documents, orders, and drafts. This allows the steno to understand context, whether it’s “Panchnama” in a trial court or a reference to a judgment by Hon’ble Justice Chandrachud.
Not to mention, the tool’s multilingual capabilities, which are a genuine breakthrough. A Bombay HC regular puts it bluntly, “Half my practice is code-switching between English legalese and Hindi. Finally, a tool that gets it.” Realistically, a lot of legal work is conducted in English, Hindi and multiple other Indian languages. In fact, several lower courts function entirely in regional Indian languages. The ingenuity of jhana steno is that it can transcribe dictation from most Indian languages, truly making it an indigenous tool for lawyers that practice in tier 2 and 3 cities and in smaller courts.
Lawyers are rarely linear thinkers. Midway through a dictation, new arguments arise or earlier paragraphs need revisiting. jhana steno’s standout feature is voice-driven editing: a simple command like, “Go back to paragraph 2 and change ‘plaintiff’ to ‘petitioner,’” triggers instant correction, without interrupting the dictation flow. This small innovation mirrors how legal minds actually work, letting drafting keep pace with thought.
jhana steno allows you to upload and maintain your templates and styles. Pause any session and apply an existing template or set of editing instructions - the steno automatically extracts your dictation contents and typesets them in your chosen template, with the requested changes. This might include prepending and appending your writ preamble and footer templates. It might entail extracting the dictated clauses and populating a readymade SLA. Or it might be a set of replacements and overall changes. Whatever your template demands, the steno is ready for the challenge.
In the pursuit of equitable access to justice, it is paramount that the efficacy of legal representation stems from the strength of legal argument and substantive knowledge, rather than the incidental advantage of superior transcription or intricate document formatting skills.
“Talent is everywhere, access shouldn’t be a privilege. We set out to democratize legal tech for every lawyer, not just those in big cities,” says Hemanth, Co-founder, jhana AI.
For the first time, dictation technology feels genuinely made for India. "It's not just about access, it’s about confidence,” adds Em McGLone, Co-founder, jhana AI. “We want every Indian lawyer/judge to draft confidently in their own voice, in their own language, without second-guessing the technology. When your words appear exactly as you intended, you can focus on the argument, not the formatting.”
Indian legal tech has been playing catch-up for too long, adapting foreign solutions to local needs. jhana steno signals a turning point, a tool that doesn’t just digitize, but truly understands the realities of Indian legal practice, this might just be the upgrade Indian law has been waiting for.
Disclaimer: This is a sponsored post from jhana steno.