The Indian court portal was built to answer one question. What is the status of a specific case. You type a CNR, you enter a captcha, you read a status page. That answers the lookup question for one matter at a time. It is the right design for a public records system that has to handle millions of one-case queries every day at predictable cost.
The next decade of Indian legaltech is not about better one case lookup. It is about a completely different question that the government portal was never built to answer. Not “what is the status of this case” but “what is happening across all the cases that affect this entity”. That second question is case intelligence, and the product class that answers it is what comes after the portal.
What case lookup does and does not do
Case lookup answers the status of a single matter, while case intelligence answers what is happening across every case tied to an entity. The government portal handles the one-case question well. The product class that comes next aggregates records, resolves entities and analyses patterns at portfolio scale, which the public portal was never designed to do.
Case lookup is the basic transaction. The user knows one matter, identifies it by CNR or case number, and asks the portal for the current status. The portal returns hearing dates, parties, advocates, the last order, the next listing. The transaction is complete in a single response.
Case lookup answers, well, a specific kind of question. Is my matter listed tomorrow. Has an order been uploaded. Who is the opposing counsel. What is the next hearing date.
Case lookup does not answer the questions that determine value at portfolio scale.
What is the full litigation exposure of this company across India.
Which advocates have argued against this opposing counsel in the past, and in which courts, and with what outcomes.
Which judge of the Allahabad High Court Lucknow Bench is statistically likely to decide cheque bounce matters in favour of the complainant.
Which of my 200 active matters are scheduled for hearing in the next seven days.
When was the last time anyone filed a writ petition naming this state department, and what was the outcome.
Are there parallel proceedings against the same party in two different courts, possibly being argued by different lawyers in unrelated firms.
What is the average disposal time for partition suits in Patna versus Pune, given similar fact patterns.
These are case intelligence questions. They sit at portfolio scale. They require entity resolution. They require cross court search. They require historical depth and pattern analysis. The one-case lookup transaction model is structurally unable to deliver them, by design and by intent.
Why the portal was not built for intelligence
Worth being honest about this. The eCourts services portal is the right design for what it is. A public records system needs to answer one case at a time, behind a captcha to prevent bulk extraction, on infrastructure sized for millions of small queries. The architecture optimises for predictable cost per query and abuse resistance.
That architecture is inconsistent with intelligence at scale. Intelligence requires bulk access. It requires structured outputs. It requires entity resolution across name spelling variations. It requires API surfaces with predictable contracts. Each of those is a different design choice than what a public records portal makes.
The right way to think about this is the same way we wrote about the broader architecture in The Operating System for Indian Law. The public portal is the foundation. The intelligence layer is built on top, by private actors who can invest in the structured aggregation work that the public layer was not built to do. The relationship is complementary. The public substrate stays cheap and abuse resistant. The intelligence layer sits on top and serves the use cases the portal cannot.
What case intelligence actually looks like
The product class is broad. Five concrete shapes.
Portfolio monitoring. A General Counsel, a senior partner, or an in house legal ops lead wants to see, in one view, every active matter against the company or for the client. With next hearing dates, recent orders, advocate assignments and risk flags. The Friday spreadsheet replaced by a Claude prompt that runs in two minutes. We documented the operational version of this in Litigation Portfolio Monitoring for General Counsel: A Claude + eCourtsIndia MCP Playbook.
Counterparty risk scoring. A bank, an insurer, a fintech, a BGV vendor wants to know the full litigation exposure of a borrower, applicant, vendor or counterparty. Across all of India, across every type of matter, with pattern recognition that flags categories of risk. This is the dataset version of the CIBIL parallel we wrote in Why India Needs a CIBIL for Litigation.
Bench and judge intelligence. A practising advocate preparing for a hearing wants to know how the bench has decided similar matters in the past, what order patterns the judge follows, and which arguments have been received well or poorly. This is the part of preparation that has historically lived in the institutional memory of senior partners.
Cause list intelligence at scale. A firm with two hundred active matters across forty courts wants its cause list prep done overnight, sorted by courtroom and item number, with adjournments flagged. The 5 AM problem solved by an MCP call. We wrote about that pattern in The 5 AM Cause List Problem.
Sector pattern analysis. A regulator, a journalist, a policy researcher, an academic wants to see what is happening at scale in a category. PMLA, Section 138, family disputes, NCLT, election petitions. The datasets we have published, including The PMLA Bail Reality Check and The Election Petition Graveyard, are early examples of this product class.
Five shapes, one substrate underneath. The same structured court data layer powers each.
Why this is a different product class from research
It is worth distinguishing case intelligence from legal research, because the established Indian legaltech category, Manupatra and SCC Online, sits in the research lane.
Legal research is about precedent. You read judgments. You build arguments around them. You file briefs that cite them. The user is preparing for a hearing or writing a memo. The product is a search tool over a curated database of appellate decisions.
Case intelligence is about ongoing matters and the entities behind them. You watch a portfolio. You spot risk. You measure pattern over time. The user is operating at portfolio scale. The product is an alerting and reporting layer over a live structured dataset of all active and disposed matters.
These are different jobs. Research can ignore the district court tier because the precedent value at that tier is limited. Intelligence cannot ignore the district court tier because the volume of activity is there. Research is read mostly. Intelligence is monitor continuously. Research is a tool the lawyer opens occasionally. Intelligence is a tool that runs in the background and surfaces what changed.
The implication is that the Indian legal AI category will see the research incumbents and the intelligence platforms as separate categories. Both can be valuable. They serve different jobs, different users and different willingness to pay surfaces.
Why the intelligence layer is the bigger category
Three reasons it is structurally larger than research.
The buyer pool is wider. Research is bought primarily by lawyers preparing for appellate work. Intelligence is bought by lawyers, by in house counsel, by banks, by insurers, by BGV firms, by RegTech vendors and by corporate diligence teams. Every entity that has to track or assess litigation exposure is in the addressable buyer pool.
The use is continuous, not occasional. Research is opened a few times per matter. Intelligence runs daily, weekly, monthly. Recurring use means recurring revenue and the SaaS economics that follow.
The data substrate compounds with use. Every entity resolved, every spelling variation captured, every new court added makes the substrate sharper for the next user. Research, by contrast, sits on a curated corpus that grows slowly.
We made the broader category math argument in India’s Legal AI Will Be a Ten Billion Dollar Category. The split between research and intelligence is one of the reasons the ten billion dollar number is conservative. The intelligence layer is the bigger half of the math.
What this means for product design
Three design implications fall out cleanly.
The product surface has to support portfolio thinking. Not just one CNR at a time but a list, a tag, an entity, a sector. Filters, client-based watchlists and update alerts.
The data layer has to be live. A weekly refresh is too slow for this product class. Daily at minimum. Sub hour on cause lists for the next working day.
The agent layer has to be addressable. Most users will not learn a new UI. They will ask in plain language through Claude, ChatGPT or a similar surface. The product has to be reachable as an MCP server so the agent can fetch the data and the user does not have to leave her existing chat tool. We laid out the architecture in MCP and the Agent Layer.
These three constraints define the product. Portfolio views, live data, agent reachable. The intelligence layer that wins is the one that holds all three to production grade.
What this means for eCourtsIndia
The next decade of Indian legaltech is not about better one case lookup. The portal does that fine. The next decade is about case intelligence at portfolio scale. Lawyers monitoring two hundred matters at once. General Counsel running weekly risk briefs. Banks scoring counterparty litigation exposure. Regulators watching sector patterns. The substrate that makes all of that possible is what we are building. Intelligence is the larger half of the category and the part that compounds.
TL;DR
- The government portal answers one-case lookup at a time. That is the right design for a public records system.
- The next product class is case intelligence at portfolio scale. Different question, different architecture, different buyer pool.
- Five concrete shapes. Portfolio monitoring, counterparty risk scoring, bench and judge intelligence, cause list at scale, sector pattern analysis.
- Intelligence is a structurally larger category than research. Wider buyer pool, recurring use, substrate that compounds.
- Product design implication. Portfolio views, live data, agent reachable through MCP.
Sources
- National Judicial Data Grid pendency dashboard, accessed May 2026
- Press Information Bureau release on eCourts Phase III
- Anthropic Model Context Protocol specification, 2024-2026
- Manupatra survey: Adoption of AI in the Indian Legal Landscape, 2025
- All India court coverage figures verified against the eCourtsIndia structured data index
Read next: The Operating System for Indian Law and Why India Needs a CIBIL for Litigation.
Frequently Asked Questions
What is the difference between case lookup and case intelligence?
Case lookup answers the status of one matter at a time. You enter a CNR or case number on the government portal and read the hearing dates, parties and last order. Case intelligence works at portfolio scale, resolving entities and searching across courts to show everything affecting a company or client. You can start a single lookup at ecourtsindia.com/search.
Why was the eCourts portal not built for intelligence at scale?
The eCourts services portal is a public records system. It answers one case at a time, sits behind a captcha to prevent bulk extraction, and runs on infrastructure sized for millions of small queries. Intelligence needs the opposite: bulk access, structured outputs, entity resolution and predictable contracts. Those design choices live in a separate layer, reachable through the eCourtsIndia API.
What are the main shapes of a case intelligence product?
The article describes five concrete shapes: portfolio monitoring, counterparty risk scoring, bench and judge intelligence, cause list intelligence at scale, and sector pattern analysis. Each one sits on the same structured court data layer. The operational version of portfolio monitoring for General Counsel is covered in our litigation portfolio monitoring playbook.
How is case intelligence different from legal research?
Legal research is about precedent. You read appellate judgments and build arguments around them, opening the tool occasionally for a hearing or memo. Case intelligence is about ongoing matters and the entities behind them. It runs continuously, monitoring portfolios and surfacing what changed across active and disposed cases. Public litigant pages list every case tied to a party at ecourtsindia.com/litigant; to get ongoing alerts on a matter, a case is mapped to a client in the dashboard, which then sends Email and WhatsApp updates.
Why is the intelligence layer a larger category than research?
Three reasons. The buyer pool is wider, reaching lawyers, in house counsel, banks, insurers, BGV firms and diligence teams. The use is continuous rather than occasional, which brings recurring revenue. And the data substrate compounds with use as more entities and courts are added. The counterparty scoring case is laid out in Why India Needs a CIBIL for Litigation.