eCourtsIndia.com launched publicly in April 2025. Twelve months later, in April 2026, we cover 37 states and union territories, 26.8 crore case records, and 29 lakh advocate profiles, delivered through a public web interface, a REST API, and the eCourts MCP. This post is an operator’s log of what it actually took to get here. We share it because we think the infrastructure layer of Indian legaltech needs more transparent operator writing, and because some of what we learned is directly useful to anyone building on top of public court data.
This is not an internal-metrics post. We do not share ARR, MRR, valuation, or fundraise details. We share the technical and operational learnings, and we stay factual about what we have built. The product is the point.
The build, in four phases
Twelve months is a short window for national court coverage in India. The work broke into four phases, roughly three months each.
| Phase | Focus | Key outputs |
|---|---|---|
| Apr to Jun 2025 | Foundation | Supreme Court and top 8 High Courts. Initial ingestion and schema. Internal testing at 1 crore records. |
| Jul to Sep 2025 | Breadth | Remaining High Courts. District coverage in 15 states. Public web UI launched. |
| Oct to Dec 2025 | Depth | District coverage completed in remaining states. NCLT, NCLAT, and key tribunals added. Advocate profile build-out. |
| Jan to Apr 2026 | Developer-first | REST API public launch. eCourts MCP public launch. Enterprise onboarding and SLA tier. |
Five things that mattered more than we expected
- Schema consistency across states. Each state presents slightly different fields, ordering, and conventions. Normalising into a single schema without losing state-specific meaning was the hardest unglamorous problem of the first six months.
- Entity resolution. The same company shows up as “ABC Pvt Ltd,” “ABC Private Limited,” “ABC (India) Pvt Ltd,” and a dozen variations. Individuals show up in every spelling combination of their names. Clean entity resolution is what separates a database from a dataset.
- Crawl stability. Public portals change shape, rate-limit unpredictably, and sometimes go offline for maintenance. Running a national crawl at daily cadence across 37 jurisdictions is a full-time infrastructure job.
- Freshness SLAs. “Daily” is not a single thing. Some courts update within hours of a hearing. Some update with two or three days of delay. Communicating realistic freshness expectations to enterprise buyers matters as much as the technical freshness itself.
- Enterprise compliance. Financial services and BGV customers require SOC 2, ISO 27001, data residency, and detailed audit logging. Building those in from the start is cheaper than retrofitting.
Five things that mattered less than we expected
- Brand. In the first year, a name that sounds like public infrastructure, which eCourtsIndia.com does, is a feature. Enterprise buyers understand the positioning faster when the name tells them what they are getting.
- Complex pricing. A straightforward tiered pricing with clear usage limits serves most buyers. Exotic pricing models are a distraction at the infrastructure layer.
- Marketing spend. We spent close to nothing on performance marketing in the first twelve months. The best inbound came from founder networks, content, and direct outreach. See our Zerodha playbook post for the reasoning.
- Deep vertical features. Targeted workflow features for a single customer category (say, litigation firms) came later. The data platform served the enterprise buyer off day one.
- Consumer UX. Our buyers are professionals and developers. They want speed and accuracy. Pretty charts did not move the needle.
Three decisions we would defend
Over twelve months, a few architectural and product decisions turned out to be more important than they seemed at the time.
- API-first, not UI-first. We invested heavily in API documentation, SDK quality, and developer onboarding from the beginning. It shortened enterprise sales cycles and attracted AI-native buyers that we did not originally scope.
- MCP early. Publishing an MCP for eCourts data in 2025 put us in front of AI-agent developers before most of the category had thought about Indian data sources. It has been one of the most effective distribution surfaces we did not expect.
- Respectful positioning. We have consistently positioned as complement to ecourts.gov.in, Indian Kanoon, and the existing ecosystem. That has earned us goodwill from practitioners and policymakers that an adversarial posture would have cost.
What we are still working on
- Deeper coverage of taluka-level courts where data release cadence is slower.
- Full-text search quality across regional-language orders.
- Judge-level analytics that respect the sensitivity of such a product.
- More mature webhook and change-stream products for enterprise customers running continuous monitoring.
What this means for eCourtsIndia
The first twelve months established the foundation. The next phase is about refining depth, enabling more integration, and continuing to support the ecosystem of products that depend on clean court data. We remain a complement to the eCourts foundation, not a replacement for it, and we remain focused on infrastructure, not destination products.
Explore the current state of eCourtsIndia.com, check the API documentation, or plug our MCP into your AI stack.
Related reading
- The Zerodha Playbook for Indian Legaltech
- The Bloomberg Terminal Analogy
- Mapping India’s Court Data Stack
Sources
- eCourtsIndia.com product telemetry and internal engineering logs, April 2025 to April 2026
- Public product launch announcements
- eCourts Services Portal, ecourts.gov.in