The District Court Lawyer Is Ninety Percent of Indian Law. Why Legaltech Keeps Missing Them.

Walk into any district court complex in India between ten and eleven in the morning, in Patna, Pune, Coimbatore or Hisar, and the scale is immediately obvious. Hundreds of advocates in black coats. Cause lists pinned outside courtrooms. Clients waiting in corridors. Stenographers carrying files. This is the working surface of Indian law. The Supreme Court hears around seventy thousand matters a year. India’s district courts hear closer to four crore. The numbers are not close.

Around eighty to ninety percent of practising Indian advocates spend their working day in this layer. The Niti Aayog estimate puts the practising bar at seventeen lakh. The Bar Council of India puts enrolment closer to twenty lakh. (The two figures count different things: enrolment, around twenty lakh, is everyone on a state roll, while the practising or verified figure of about seventeen lakh is those actually appearing in court.) The vast majority of them argue at the trial court tier. And yet, the Indian legaltech category has spent two decades building for the other ten percent.

This post is about why that mismatch happened, what it costs the country, and why the next decade of Indian legal AI will be defined by whoever finally builds the product the district lawyer can use.

Where the bar actually works

About 80 to 90 percent of India’s roughly 17 lakh practising advocates work at the district court tier, where the courts hear close to four crore matters a year. This is where most of Indian law actually happens, yet legaltech has spent two decades building for the appellate ten percent.

The structure of Indian legal practice is shaped by a pyramid. At the top, around three hundred and fifty senior advocates and a few thousand others argue regularly at the Supreme Court. Below them, a few tens of thousands argue at the twenty five High Courts and their forty plus benches. Below that sit seven hundred plus district court complexes, each with their own courts, civil judges, magistrates, sessions courts, special courts and family courts. That layer is where most of the bar makes its living.

The work at the district tier is different from what people imagine when they think about Indian law. It is criminal complaints under Section 138 of the Negotiable Instruments Act. It is property disputes that span generations. It is family court matters under the Hindu Marriage Act, the Special Marriage Act and the Indian Divorce Act. It is motor accident claims at the MACT. It is consumer disputes at the District Forum. It is bail applications, appeals from magistrates, writ petitions against state authorities, and the long, grinding civil suits that run for two or three decades.

The data scale tracks. The eCourtsIndia structured index, which sits on top of the Case Information Software that the eCourts Mission Mode Project deployed across the district tier, sees the overwhelming majority of its records at this layer. Section 138 cases alone account for over forty three lakh pending matters nationwide as of early 2026, per the Ministry of Law and Justice estimate we covered in Cheque Bounce: India’s Silent Litigation Crisis. Family property disputes account for another four lakh files between partition, succession, probate and mutation, as of 2026, per the dataset we published in The Indian Family Property Dispute Index 2026. (These pendency figures refresh as courts update their records.) The district tier is not the smaller cousin of the appellate tier. It is the bulk of Indian law.

Why legaltech has missed this layer

The history is unflattering and worth being honest about. The Indian legaltech category has, for two decades, built tools that price out the trial court advocate by design.

Manupatra and SCC Online, the two largest legal research incumbents, charge between fifteen thousand and fifty thousand rupees per seat per year, depending on the tier. Their core product is judgment search, concentrated on Supreme Court and High Court material. District court coverage has historically been thin or non existent. For a senior counsel at a Delhi or Mumbai firm, that price and that coverage is rational. For a sessions court advocate in Allahabad earning a few lakhs a year, neither is reachable.

Westlaw and LexisNexis, the global research incumbents, have no real India product. The few products they do offer are positioned at the global research tier and price accordingly. The contract management category, the CLM space, is built for in house teams at corporates, not the district bar. The dispute resolution and ODR category is built for fintech and consumer companies, not the district bar.

The result is a product class that fits the top ten percent of the market and ignores the eighty to ninety percent underneath. The district advocate has, in effect, been told by the legaltech industry that her workflow is not worth building for. She has responded the way any rational professional would. She has not bought the tools. She has improvised. She has built her own systems in WhatsApp, in physical files, in junior memory.

This is not a complaint about the incumbents. They built what they could afford to build given the data and the market they understood. It is a description of the gap.

Why the gap was structurally hard to close

The reason no one built for the district lawyer is not lack of will. It is that doing so requires solving three hard problems at once that no incumbent had the architecture to solve.

Data coverage. District court data is fragmented across seven hundred plus complexes, each on its own portal flavour. Records are often scanned PDFs in vernacular. The data simply did not exist in structured form at the scale required for a useful product. Manupatra and SCC Online built their corpora on the Supreme Court and High Court reporter system, which is a curated, English language, post hearing publication track. That same track does not exist at the district tier. You cannot put a research product on top of records that do not exist in your database.

Language. The trial court advocate works in the regional language of her bench. Hindi in UP, Bihar, Madhya Pradesh, Rajasthan, Haryana, Delhi at the lower tier. Marathi in Maharashtra. Tamil in Tamil Nadu. Telugu in Andhra Pradesh and Telangana. Bengali in West Bengal. Kannada in Karnataka. Gujarati in Gujarat. Odia in Odisha. Punjabi in Punjab. An English first product class never landed.

Price point. Even a fifteen thousand rupee per year seat is a serious purchase for an advocate who clears two to four lakh rupees a year. The price has to start at hundreds of rupees per month and earn its way up. Western pricing models do not apply. Indian SaaS pricing models for B2B do not apply. The product has to be priced at the bar’s actual willingness to pay.

Solving the data problem required a structured data layer at national scale that nobody had built. Solving the language problem required vernacular OCR at production grade. Solving the price problem required a software architecture that did not depend on the high margin enterprise sales motion. The three problems compounded. The product that wins the district tier has to clear all three bars at once.

What changed in the past five years

Three things happened in parallel that finally make the district court product economically feasible.

Phase II of the eCourts Mission Mode Project digitised the district tier. More than twenty nine thousand court establishments now sit on Case Information Software. The underlying digital records exist. The portals are imperfect, but the data is there. Phase III, with its Rs. 7,210 crore outlay between 2023 and 2027 per the Press Information Bureau, will fund the next leg of integration, OCR and AI capability inside the public infrastructure. The public foundation is in place.

Frontier AI made vernacular legal work feasible. Generative AI’s ability to read and produce high quality output in Hindi, Marathi, Tamil and the other major Indian languages has improved dramatically since 2023. Translation and summarisation that would have required a junior associate or a court translator can now be done by Claude or GPT in seconds. The unit economics of vernacular legal work changed.

Distribution moved to the phone. The district advocate does not own a workstation. She owns a phone, and increasingly a tablet. The product surface that wins this market is mobile first and supports offline drafting. The Play Store and the App Store now reach the bar in a way that the law firm intranet never did.

Stack these three shifts together and the district tier moves from structurally unbuildable to structurally inevitable. The first team that ships a vernacular ready, mobile first, AI enabled product on top of the public data layer at the right price point will define the category.

What the district lawyer’s product actually has to do

The product class is different from the appellate research product class. Worth being specific about what it needs.

Unified case management. All matters, across all courts the advocate practises in, in one view. Next hearing, last order, opposing counsel, judge, client. The Five AM cause list problem we wrote about in The 5 AM Cause List Problem, and How to Fix It With One MCP Call is the daily killer. Solve it and the advocate cannot go back.

AI drafting in vernacular. Replies, written submissions, applications, complaints, all in the language the bench reads them in. Generated against the matter context, the bench’s prior orders on similar matters, and the local procedural conventions. This is the part where the AI Lawyer becomes a tangible time saver instead of a novelty.

Smart alerts. When the court uploads an order. When a hearing is moved. When the status changes on a matter. The advocate maps the case to a client, opts in, and the updates land by email and WhatsApp, near real time, instead of forcing a portal login every morning.

Cause list reading and prep. Auto sorted by courtroom, item number, board number. Flags adjournments and arguments. Pulls the latest order on each listed matter so the advocate walks in prepared.

Document vault and search. Every PDF, every pleading, every order, every annexure, indexed and searchable. The advocate can ask, in plain language, what the respondent argued in the last hearing. The AI Clerk pattern, in production.

Lawyer, judge and litigant pages. The advocate can pull up her opposing counsel’s public case history in the same bench, the judge’s past orders, and her client’s other matters, all drawn from the structured court record. A lawyer can claim and edit her own profile.

This is one product, one subscription, one phone. The advocate’s working day inside a single surface. Pricing has to start low, scale with usage and feature depth, and pay for itself in time saved within the first month.

We covered an early version of how this plays out for a real practising advocate in One Advocate, Nine Courts, One Claude Window. The pattern works.

Why this is the largest legaltech opportunity in the country

The math is straightforward. Seventeen lakh practising lawyers. Eighty to ninety percent of them at the district tier. The bar has been almost entirely unmonetised by the legaltech industry for two decades. The product class that finally serves this layer does not need to displace an incumbent. It just needs to exist and be reachable.

Compare with the Western markets that get most of the venture funding attention. The American Bar Association reports roughly thirteen to fourteen lakh active lawyers in the United States, of whom maybe three hundred thousand are at the AmLaw 1000 tier that Harvey and Legora are built for. The Indian district tier is itself larger than the entire American legal market that drove the Harvey valuation. The number is hard to absorb until you say it out loud.

The implications are not theoretical. If even ten percent of the Indian district bar pays a modest subscription for a workflow product, that is one point seven lakh paying users. Apply a mature average revenue of around twenty four thousand rupees per year and you arrive at four hundred crore rupees of annual subscription revenue from the lawyer surface alone. Apply Harvey style revenue multiples to that and the lawyer business alone is worth two to three billion dollars. We laid out the full sizing logic in India’s Legal AI Will Be a Ten Billion Dollar Category.

The district tier is the volume layer. The value of the category does not come from a handful of high value enterprise contracts. It comes from a few hundred thousand advocates running their working day on the same product. That is a SaaS shape that India has only seen a handful of times. Zerodha for stock broking. Khatabook and OkCredit for small business ledgers. PhonePe and Paytm for payments. The shape repeats. Captive user base, daily workflow, replace an incumbent that was government infrastructure, price at a fraction of the Western comparable, build the moat through distribution and habit.

The data layer is the unlock

None of this is possible without the structured court data layer underneath. Every feature listed above requires the same foundation. Unified case search needs cross court entity resolution. Smart notifications need live freshness. AI drafting needs the bench’s order corpus. Cause list reading needs the daily refresh. Lawyer and judge profiles need the cross linked index.

This is why the question of who serves the district lawyer turns into the question of who built the data layer. The two are inseparable. The product that wins is the product whose data layer was built to support it. There is no path through scraping or partnerships or wrapping a foreign AI in an Indian skin. The data has to be owned, structured and refreshed by the team that ships the product.

We have written about why the data layer is the hardest part of Indian legaltech in The Operating System for Indian Law. That post lays out the broader architecture. This post is about the user the architecture is built for.

The district court lawyer has waited two decades for a product that fits her actual working day. The foundations are now in place. The first team that ships at scale defines the category for the rest of the decade.

What this means for eCourtsIndia

Most Indian law, eighty to ninety percent of it, happens at the district tier. The bar at that tier has been ignored by every previous wave of Indian legaltech. The structural reasons are clear. Data coverage, language, price points. Those reasons are no longer binding. The next product class will be vernacular ready, mobile first, AI enabled and priced for the trial court advocate, sitting on a structured court data layer that the team that ships it also owns. That is the product we are building.

TL;DR

• The Indian practising bar is around 17 lakh, with 80 to 90 percent of advocates working at the district court tier per Bar Council and Niti Aayog estimates.
• Legaltech incumbents like Manupatra and SCC Online price at Rs. 15,000 to Rs. 50,000 per seat per year, focused on Supreme Court and High Court research. District tier coverage is thin.
• The structural blockers were data coverage, vernacular language and price points. All three are now solvable for the first time, thanks to eCourts Phase II and III, frontier AI vernacular capability and mobile first distribution.
• The product class the district advocate needs is a single workflow surface. Unified case management, AI drafting in vernacular, smart notifications, cause list reading, document vault and cross linked profiles.
• This is the largest underserved professional buyer pool in Indian SaaS. The product that wins it is built on a structured court data layer owned by the same team that ships the product.

---

Sources

• Niti Aayog estimate of 17 lakh practising lawyers in India
• Bar Council of India enrolment data, approximately 20 lakh
• Press Information Bureau release on eCourts Phase III outlay, Rs. 7,210 crore between 2023 and 2027
• Ministry of Law and Justice data on Section 138 pendency, 43,05,932 cases (as of early 2026)
• American Bar Association lawyer count for the United States, 2025
• Reuters reporting on Harvey AI USD 11 billion valuation, March 2026
• All India court coverage and case type counts verified against the eCourtsIndia structured data index

Read next: The Operating System for Indian Law and India’s Legal AI Will Be a Ten Billion Dollar Category.

Frequently Asked Questions