Langdell predicts the ruling.
Case research that predicts outcomes from your facts. Due diligence that writes its own report. Matter management that reads every order. All on one graph of Indian law, every answer cited to the paragraph.
Langdell simulates the hearing a thousand times, grounded only in documented rulings.
Explore case research →Each query is routed to keyword, phrase, semantic and structural retrieval at once, then re-ranked by court hierarchy, recency and treatment.
Select a passage from your draft. Langdell ranks holdings by how strongly they contradict it and how much authority they carry, and explains why each one cuts against you. The best adverse case is rarely the nearest neighbour.
A judgment with several holdings gets a score for each one, not a single red or green light.
Where coordinate benches have split and neither has been set aside, you see it. Search mode finds every split on a topic.
So Langdell is a pipeline, not a prompt. It answers only from holdings it has retrieved and verified, and every judgment it reads is kept at the source, the original PDF untouched.
Every quote is matched to its source text. Anything that doesn't match is removed and flagged.
Each case is checked against the database before it renders. Unverifiable citations are blocked.
"High, 3 corroborating judgments" or "low, single source, verify manually", next to the answer.
When retrieval is weak, Langdell says so instead of guessing.
Every statement opens to the paragraph it came from.
Full text, holdings and fact pattern are indexed apart, so a fact query never falls back to word matching.
Your research task, broken into the questions a court would ask.
Each judgment read for facts, issues, holding and ratio.
Every authority followed forward to the cases that relied on, distinguished or overruled it.
NCLT, NCLAT and tribunal listings checked for anything new on your matters.
A structured report where each point links to its paragraph.
Escrow releases, earn-out milestones, deferred consideration and non-compete periods are pulled from the signed SPA. Indemnity and warranty survival periods run on countdowns, synced to Outlook or Google Calendar.
Mark unpublished price-sensitive documents. Every access is logged by person, date and document, ready for the compliance officer.
A restricted sub-room for pricing and customer terms, to reduce gun-jumping risk before CCI approval.
Release documents in phases: phase 1 to every bidder, phase 2 to the shortlist.
Every issue framed and every material averment is listed, tagged to the party who must prove it, and matched to its evidence and authorities. Written submissions are drafted from the links, each point citing the exhibit and the judgment paragraph.
Every new party checked against your existing matters and clients, with the result logged.
Restrict a matter to named team members, so sensitive disputes stay with the people working on them.
Mark privileged documents. Every access is logged by person, date and document.
Thresholds, section numbers and procedural rules change often. Every rule is stored as versioned, counsel-reviewed data with an "as of" date shown in the product, never hard-coded.
Land records, older corporate records, district court orders and government orders, read with a confidence score on every page.
Registered charges, filing defaults and director issues read from MCA21 and raised as risk flags automatically.
FEMA and RBI approvals and FC-GPR reporting checked against the transaction, not left to a checklist.
Stamp duty checked on each share transfer in the target's history.
GST returns checked for every month in the review period, with each gap raised as a finding.
A dispute running across NCLT, a High Court and the Supreme Court is linked as a single matter family.
The judgments Langdell reads, the companies it scans and the matters it tracks sit on the same legal graph. What one product finds, the others already know.
A tribunal matter found in diligence opens with its judgments already analysed.
Matters, drafts and deal documents are some of the most sensitive data a firm holds. This is how Langdell is being built to protect them.
Court data and client documents are designed to stay on India-hosted infrastructure, in line with the DPDP Act.
At rest and in transit, including deal documents that share infrastructure with case search.
Each search, answer and agent action is attributable, so you can show exactly what happened and when.
SSO, multi-factor sign-in and roles for associates, partners and admins. Buyer and seller teams stay apart.
Agents never send an email, write to your calendar or file anything without your confirmation.
Victims' names in sexual-offence and juvenile matters are masked, as Indian law requires.
For law students, judges and working legal professionals. Pick the products you want early access to.
Langdell reads every judgment the way you would, matches your facts to the facts courts actually decided on, and tells you how the bench is likely to rule. Each estimate comes with its range, the precedents behind it, and the exact paragraph it relied on.
Open to students, judges and working professionals. One email when your seat opens.
Why we built it
Legal databases built their moat on how many judgments they hold. None built it on reasoning. Langdell reads a judgment the way a lawyer does, facts, issues, holding, ratio, and then tells you where the law stands *on your facts*, with the paragraph to prove it.
Langdell syncs the Supreme Court archive from 1950 and 25 High Courts daily, keeps the original PDF untouched, and never stores the same judgment twice.
Keyword search finds judgments that use your words. Langdell extracts the fact pattern of your matter, parties by role, timeline, delay, posture, quantum, conduct, and searches a separate index of the facts courts decided on. Then it tells you which differences the holding actually turned on.
Facts are embedded apart from full text and holdings, so a fact query never falls back to word matching.
A difference counts as material only if it appears in the holding, not just in the recital of facts.
Aligned facts, divergent facts, and how opposing counsel would use each difference against you. Every line cites both documents.
Langdell acts, it doesn't only answer. Every step, source read and confidence level is visible, and nothing leaves the platform without your confirmation.
Upload pleadings and prior orders. Langdell builds the case graph, retrieves the closest precedents, and simulates the hearing across the factors that are genuinely uncertain.
Illustrative. Decision support, not legal advice.
Type a proposition or select a passage from your draft. Langdell ranks holdings by how strongly they contradict it and how much authority they carry. The best adverse case is rarely the nearest neighbour.
“Limitation runs from the date the award is received, not the date it is signed.”
Langdell reads your brief and runs four checks on each authority. It flags a fabricated case as fabricated, with no softer wording.
A single red or green light can't describe a judgment with several holdings. Langdell scores each holding on its own, and flags where coordinate benches have split and neither has been set aside.
Give Langdell a research task and it works through it step by step: searching, reading, checking treatment, and drafting a structured report where each point links to its paragraph.
Langdell routes each query to keyword, phrase, semantic and structural retrieval, then re-ranks by court hierarchy, recency and treatment. Paste a citation to jump straight to it.
Can an operational creditor file under § 9 IBC when the debtor raised a dispute only after the demand notice?
6 judgments address this. 4 support your position: a dispute raised for the first time after the demand notice does not bar admission ¶ 42. The other 2 turn on correspondence that predates the notice ¶ 18–21.
The model only answers from retrieved, verified holdings.
Every quote is matched to its source text. Anything that doesn't match is removed and flagged.
Each case is checked against the database before it renders. Unverifiable citations are blocked.
“High, 3 corroborating judgments” or “low, single source, verify manually”.
When retrieval is weak, Langdell says so instead of guessing.
Every statement opens to the paragraph it came from.
| Area | Legacy legal databases | Langdell |
|---|---|---|
| Search | Keywords and metadata, with a thin AI layer | Lexical, semantic, structural and fact-pattern retrieval, fused and re-ranked |
| Citation network | A static “cited by” list | A typed graph resolved to the paragraph: followed, distinguished, overruled |
| Case status | One flag per judgment | Status per holding, with splits between benches flagged |
| Prediction | None | Grounded outcome simulation with a range and the precedents behind it |
| Hallucination control | Not applicable, mostly non-generative | Verification pipeline, confidence scoring, source trace on every line |
| Agentic action | None | Research briefs, docket monitoring, deadlines, draft audit |
| Where you work | A separate website | Word, Google Docs and Outlook add-ins |
For law students, judges and working legal professionals. It takes under a minute.
Langdell is designed as a diligence engine for Indian deals. Agents pull the records, read the data room and draft cited findings. Your counsel, chartered accountants and valuers make the calls and sign what others rely on.
Why we built it
A share purchase, an insolvency bid, a lender's check and a new vendor all ask the same question: who is this counterparty, and what could go wrong? So Langdell runs *one engine* and switches on the modes each deal needs.
A template sets which modes run at full depth and which run at red-flag depth. You can change any of them.
Before a data room exists, screen any company.
Or keep it running: portfolio monitoring re-scans your companies and alerts you to new filings.
Every finding shows the documents, rule version and steps behind it. Anything outside the checklist is labelled as unrequested and waits for a person's materiality call.
Run firm playbooks and upload your template, numbering and qualifications. Full-scope, limited-scope, vendor and lender reports.
A matrix lists each SPA warranty with its qualifiers. Paper checks compare the term sheet with the SPA, defined terms across documents, and the executed copy with the agreed draft.
Add columns in plain English and keep the verbatim clause beside each summary. It reads Excel with cell-level citations, PST mailboxes, password-protected files you authorise, and scanned Hindi and regional-language records. If a contract mentions an amendment that isn't in the room, Langdell raises it.
Planned: virtual data room connectors, MCP and agent-to-agent access, and market deal-point benchmarks.
Match the request list to documents already in the room. Duplicate questions are spotted and routed to the right workstream, and answers need a person's approval. Bidder rooms open in phases, first to all bidders, then the shortlist.
Engagement signals are labelled as heuristic, not fact.
Each change-of-control clause becomes a consent item with a pre-filled letter. E-stamp and e-sign steps are tracked against each instrument.
Behaviour monitors are designed to pause a run on a deviation, or on prompt injection hidden in data-room text. Memory is viewable and editable, and designed never to cross an ethical wall.
Langdell builds the evidence and writes the first draft. Opinions stay locked until the person named here signs.
Use your own counsel, or a panel you curate, with no ranking, ratings or marketing. Langdell's software fee is billed separately and it takes no share of an advocate's fee. In-house teams can run red-flag and vendor checks themselves where no third party will rely on the result.
Raised as risk flags automatically.
Thresholds and section numbers change often. Every rule is stored as versioned, counsel-reviewed data with an “as of” date shown in the product, never hard-coded.
Buy-side and sell-side workspaces with Chinese walls between them.
A branded space for clients and outside counsel, with access set per engagement.
Severity-rated issues rolled up into one aggregate score for the deal.
Milestones for each template and a deal health indicator.
Live links to the target's matters in tribunals and courts.
Edit in Word without losing the links back to the evidence.
No automated output changes a risk score without a person accepting it.
Nothing negotiates for you. Redlines and letters are sent by a person.
Background checks start only after the person's consent is on file.
Langdell is pre-launch. This page describes the planned product. Deal names, figures and scores are illustrative.
Set a task in the evening. Come back to a brief you can check, line by line.
Your task becomes a written list of issues and the forums to search, so you can see how it understood you.
Results are then re-ranked by court hierarchy, recency and how later courts treated them.
Each hop is filtered for relevance, and every holding is marked as followed, distinguished, overruled or doubted.
Citations are checked against the case database too. What can't be verified is stripped and flagged, so you see why it left.
Ratio is separated from obiter, and each point opens to the paragraph it came from.
Too little authority, and the question stays open.
Two benches split, and you read both.
Confidence rates the sources it found, not your chances.
It prepares the work, then it stops and waits for a person.
No. It finds, reads and checks faster, and a lawyer verifies the result. Judgement stays with you.
Langdell is designed to keep court data and client documents on India-hosted infrastructure, encrypted at rest and in transit.
We will share pricing with early-access teams before general release.
Langdell is pre-launch. This page describes the planned product. Case names, dates and figures are illustrative.
The search and verification behind each step.
Lands inMatter management →Where watched dockets and deadlines end up.
Join the waitlist for early access to the agent.
Most case trackers store dates. Langdell runs the matter: it pulls the docket, reads every order, computes limitation and compliance deadlines, and turns every hearing, direction and research note into one record that flows into client reports, filings and appeals.
Why we built it
Case trackers remember dates. They don't read the orders. Langdell reads every order the day it lands, turns each direction into a task, and *keeps the record for you.*
Paste a CNR or case number and Langdell pulls the case status, hearing history, orders and parties, then returns a one-page matter snapshot with the next date. Every item is source-linked with a “last synced” time.
Or keep it running: docket monitoring watches your matters and alerts you to every new listing and order.
One-click client status report or partner brief, by stage: pleadings, interim relief, evidence, arguments, orders. Each entry cites its order and paragraph; unsupported sentences are blocked.
Your own formats for client updates, board notes, insurer reports and litigation-hold summaries.
Associates draft, partners approve. A new order after sign-off marks affected entries “stale”. Export to Word or PDF.
Every issue framed and every material averment in the pleadings listed, tagged to the party who must prove it.
Issues against linked evidence and authorities, each with a status:
Draft written submissions from linked evidence and authorities, each point citing the exhibit and the judgment paragraph.
Each direction (“reply within four weeks”, “deposit 50% of the award”) becomes a task with an owner, with the draft opening in the editor.
Every task appears on the matter timeline. An outstanding direction before the next date turns matter health amber.
Limitation for suits, appeals and applications computed from the facts and the date of order, with the working shown and time for certified copies excluded.
Countdowns from the date of order or receipt, with alerts before expiry.
Court-ordered deposits, undertakings and stays tracked to their dates, synced to Outlook or Google Calendar.
Every new party checked against your existing matters and clients before the matter opens, with the result logged.
Restrict a matter to named team members, so sensitive disputes stay visible only to the people working on them.
Mark privileged documents; every access is logged by person, date and document.
Directions, next dates, findings and reliefs extracted from every order and pleading, with a human sign-off on every result.
Hindi and regional-language OCR for district court and tribunal orders, with a confidence score per page.
One dispute across NCLT, the High Court and the Supreme Court linked as a single matter family.
Case files sorted into paper-book order and numbered, with a review queue for low-confidence files.
| Check | What it does | Source | Output |
|---|---|---|---|
| CNR docket pull | Enter a CNR or case number; fetch status, hearing history, orders, parties and advocates | eCourts and court websites | Auto-filled matter record; new orders raised as tasks |
| Cause-list watch | Daily cause lists of the Supreme Court, High Courts, NCLT and NCLAT checked for your matters | Published cause lists | An alert and calendar entry when a matter is listed |
| Limitation calculator | Limitation periods by proceeding, with exclusions such as time taken for certified copies | Rules engine, curated by counsel | A deadline with the working shown |
| Court fee calculator | Court fee by state, court and value of the claim | State-wise fee schedules | Estimate in the filing checklist |
| Authority check | Judgments cited in your pleadings checked for existence, pin-cite, support and current treatment | Langdell case research | Flags on the matter before filing |
Limitation periods, fee schedules and procedural rules change. Every rule is stored as versioned, counsel-reviewed data with an “as of” date shown in the product, never hard-coded.
Parties, counsel, forum, bench and stage, with every connected matter.
Pleadings, orders and evidence with version history.
Authorities saved from case research stay linked to the matter.
Hearings, directions and deadlines with a matter health indicator.
Associate, partner and admin roles in firm-level workspaces.
Edit in Word without losing links; dates sync to Outlook and Google.
Tell us who you are and we'll write to you when your access opens.
| Name | Type | Organisation | Interested in | Joined |
|---|
Langdell is built by Dreambots Legal Tech. We're building agents for Indian legal practice that read the law the way lawyers do, and cite every answer to the paragraph it came from.
Christopher Columbus Langdell made the case the unit of legal study. We think the case, read closely and in context, is still the unit that matters. Langdell reads each judgment as a set of facts, issues, holdings and an order, then connects every holding to the cases that followed, distinguished or overruled it.
Hybrid search across the Supreme Court and 25 High Courts, fact-pattern matching, outcome prediction and a citation check for your drafts.
Due Diligence agentFrom exposure scan to closingA deal workspace with India-specific corporate, tax and regulatory checks, from target exposure scans to the diligence report.
Matter managementEvery hearing, order and deadlineA matter workspace that pulls dockets, tracks hearings and limitation, and drafts the client report from the record.
Your matters, drafts and deal documents are some of the most sensitive data a firm holds. Here is how we're designing Langdell to protect them.
Court data and client documents are designed to stay on India-hosted infrastructure, in line with the DPDP Act.
Data is encrypted at rest and in transit, including deal documents that share infrastructure with case search.
Each search, generated answer and agent action is logged and attributable, so you can show exactly what happened and when.
SSO, multi-factor sign-in and roles for associates, partners and admins. Buyer and seller teams stay behind separate walls.
Agents never send an email, write to your calendar or file anything without your explicit confirmation.
Names of victims in sexual-offence and juvenile matters are masked as Indian law requires.
We're a small team working on hard problems in legal reasoning, retrieval and verification. We don't have formal openings listed yet. If your work fits one of these areas, write to us at info@langdell.in.
Holding extraction, stance models, citation resolution and fact-pattern search over millions of judgments.
Ingestion pipelines, OCR for Hindi and regional scripts, indexing and the citation graph.
Advocates and researchers who build gold-standard test sets and check what the model gets wrong.
Interfaces that make agent reasoning, confidence and sources easy to read at a glance.
Questions about Langdell, partnerships, press or careers. Send us a note and we'll reply by email.
The policies that govern this site, the waitlist and the private beta, in plain language. If anything here is unclear, write to us and we'll explain it.
What personal data we collect, why, who processes it, and your rights under the DPDP Act.
AgreementTerms of use →The rules for using this site, joining the waitlist and taking part in the beta.
PolicyCookie policy →What we store in your browser, and why this site needs no cookie banner today.
CommitmentsAI principles →How Langdell's agents cite, show uncertainty, keep you in control and treat your data.
SecurityResponsible disclosure →How to report a security vulnerability, and what we promise in return.
StatementAccessibility →The standard we build to, what we know still needs work, and how to tell us.
TransparencySubprocessors →The providers that process personal data for us, what they do, and where the data is stored.
Your rightsMake a data request →Access, correct or erase your data, or withdraw consent. Section 11 of the privacy policy explains how.
Last updated 4 November 2026. This policy explains what personal data Dreambots Legal Tech Pvt. Ltd. collects through the Langdell website, why we collect it, who we share it with, how long we keep it, and the rights you have over it.
Langdell is built for advocates, judges, in-house teams and law students, people whose work depends on confidentiality. We hold ourselves to the same standard. In practice that means:
We've written this policy to be read, not skimmed past. If anything in it is unclear, write to us at info@langdell.in and we'll explain it.
Langdell is built and operated by Dreambots Legal Tech Pvt. Ltd., a company incorporated in India with its registered office at 4440, 1st Floor, Gali No. 55, Regarpura, Karol Bagh, Delhi 110005 ("Dreambots", "we", "us" or "our").
For the personal data described in this policy, we are the Data Fiduciary under the Digital Personal Data Protection Act, 2023 and the rules made under it (together, the "DPDP Act"). Where they still apply, we also follow section 43A of the Information Technology Act, 2000 and the Information Technology (Reasonable Security Practices and Procedures and Sensitive Personal Data or Information) Rules, 2011.
This policy covers:
It doesn't cover the Langdell product itself. When the private beta opens, the matters, documents, searches and other content you or your organisation place in Langdell will be governed by a separate customer agreement and data processing terms (see section 17).
We collect personal data only when you choose to give it to us through a form or by email.
| When | What we collect | Why |
|---|---|---|
| Joining the waitlist | Full name; email address; whether you are a student, judge or working professional; your institution, court or organisation (optional); the Langdell products you're interested in; the date you joined. | To hold your place, contact you when access opens, prioritise invitations by user group and product, and plan the beta. |
| Using the contact form | Full name; email address; the topic you choose (general, partnerships, press or careers); your message. | To read and reply to your message and keep a record of our correspondence. |
| Emailing us directly | Your email address, name, signature details and anything you include in the message or attachments. | To reply to you, and where relevant to consider a partnership, press request or job application. |
| Applying for a role | Your CV, cover note, links to your work and any other details you choose to send. | To assess your application and contact you about it. |
Please don't send us confidential client information. The Site isn't designed to receive privileged material, case files, details of ongoing matters or sensitive personal data such as health, financial or biometric information. If you send it to us by mistake, tell us and we'll delete it.
Like every website, the Site can't be delivered to your browser without some technical data changing hands. We keep this to a minimum and don't use it to build a profile of you.
| Data | Who receives it | Why |
|---|---|---|
| IP address, browser type and version, operating system, the page requested, the referring page, and the date and time of the request | Cloudflare, our website host | To deliver pages to you, protect the Site from attacks and abuse, and diagnose faults. |
| The page visited, how quickly it loaded and how steady its layout stayed, measured by a small script in your browser | Cloudflare Web Analytics | To see which pages are visited and fix slow ones. The script sets no cookies and stores nothing on your device. |
| A note in your browser's local storage that you've joined the waitlist | Stays on your device; we can't read it remotely | So the Site shows that you're already on the waitlist instead of asking again. |
| A short-lived security cookie, only if unusual traffic is detected | Cloudflare | To tell people from automated bots. See our cookie policy. |
We don't use session recording, fingerprinting, advertising cookies or any technology that tracks you across other websites.
We use personal data only for these purposes:
Our grounds for processing. We process data you submit through our forms on the basis of your consent, which you give when you submit the form after being shown this policy. Where you voluntarily send us information for a specific purpose, such as an email asking a question, we use it for that purpose as the DPDP Act permits. We process technical data to keep the Site secure, to measure how it performs, and to meet legal requirements.
We don't use your personal data to make automated decisions that have legal or similarly significant effects on you. We don't use waitlist or contact data to train AI models.
We don't sell, rent or trade personal data. We share it only in these situations:
If you join the waitlist, we'll send you emails about your place on the list and your access to the beta. These are service emails, part of the waitlist itself. We'll send occasional product updates only if you've asked for them, and every update will contain a way to stop receiving them.
We don't share your email address with anyone for their marketing, and we don't use advertising networks to target you based on your sign-up.
Everything you submit through our forms is stored in a database hosted by Supabase on servers in Mumbai, India.
Cloudflare operates a global network. When your browser loads a page, the technical data described in section 4 may be handled at the network location nearest to you, which may be outside India. The DPDP Act permits transfers outside India except to countries the Central Government has restricted, and we'll stop any transfer to a country it restricts.
We keep personal data only for as long as we need it for the purpose we collected it for, or as long as the law requires. When that ends we erase it, or anonymise it so it can no longer identify you.
| Data | How long we keep it |
|---|---|
| Waitlist details | Until you're given beta access and for 12 months after that, or until you ask us to remove you, whichever is sooner. If the waitlist closes without an invitation, we erase your details within 12 months of it closing. |
| Contact form messages and emails | For as long as we need to deal with your query, and up to 24 months afterwards in case you follow up. |
| Job applications | Up to 12 months after the role is filled, unless you ask us to keep your details for future roles or to delete them sooner. |
| Security and server logs | Up to one year, or longer where the law requires us to keep them. |
| Visit and page speed measurements | Available to us in Cloudflare's dashboard for six months. |
| Records we must keep by law | For the period the relevant law requires, for example records needed to establish or defend legal claims. |
Where the DPDP Act requires it, we'll tell you at least 48 hours before erasing your data for inactivity, so you can let us know if you'd like us to keep it.
We use reasonable security safeguards, appropriate to the data we hold, to prevent personal data breaches. They include:
No method of transmission or storage is perfectly secure, and we can't guarantee absolute security. If you have reason to believe your interaction with us is no longer secure, please contact us straight away.
Under the DPDP Act you have the right to:
How to make a request. Email info@langdell.in with the subject line "Data request", from the email address you used with us. Tell us which right you want to exercise. To protect your data, we may ask you to confirm your identity before acting on a request, and we'll use the information you give us only for that purpose.
We'll acknowledge your request promptly and respond within 30 days. We don't charge a fee. If we can't do what you've asked, for example because the law requires us to keep certain records, we'll explain why.
If you're not satisfied. You can escalate to our Grievance Officer. If you remain unsatisfied after using our grievance process, you can complain to the Data Protection Board of India.
The DPDP Act also places some duties on people exercising their rights. When dealing with us, you must not impersonate someone else, must not hide material information when giving us your details, must not file false or frivolous complaints, and must give only verifiably authentic information when asking us to correct or update your data.
We don't set cookies. We use your browser's local storage for one functional purpose, and our host may set a strictly necessary security cookie. Our cookie policy lists each item, what it does and how long it lasts.
The Site, the waitlist and the beta are intended for people aged 18 or over. Law students under 18 shouldn't join the waitlist. We don't knowingly collect personal data from children, and we don't track, monitor behaviour or target advertising at children. If you believe someone under 18 has given us their details, write to us and we'll erase them.
Where a person with a disability has a lawful guardian, the guardian may exercise rights under this policy on their behalf, and we'll take reasonable steps to verify the guardianship.
The Site may link to websites we don't operate, such as court portals or news articles. We aren't responsible for their content or privacy practices, and this policy doesn't apply to them. Please read their privacy policies before giving them your personal data.
If a personal data breach occurs, we'll act immediately to contain it and limit its effects. We'll inform affected individuals without delay, describing the breach, its likely consequences, what we've done about it, and steps you can take to protect yourself. We'll also report the breach to the Data Protection Board of India in the manner and time the DPDP Act requires, and to CERT-In where a cyber security incident must be reported to it.
When the beta opens, law firms, legal departments and other organisations that subscribe to Langdell will place matters, documents, deal data and searches in it. For that content, the subscribing organisation will usually be the Data Fiduciary and we'll act as its Data Processor. That processing will be governed by our customer agreement and data processing terms, not by this policy. These will include our commitments that customer content is stored in India, kept confidential, and not used to train models that serve other customers. We'll publish them before the beta opens.
We may update this policy as the Site, the law or our practices change. When we do, we'll change the "last updated" date at the top. If a change materially affects how we use personal data you've already given us, we'll email you before it takes effect and, where required, ask for your consent again. Earlier versions are available on request.
If you have a question, a request or a complaint about how we handle personal data, contact our Grievance Officer:
| Grievance Officer | Avisek Sil |
| Company | Dreambots Legal Tech Pvt. Ltd. |
| info@langdell.in (subject line "Grievance") | |
| Post | 4440, 1st Floor, Gali No. 55, Regarpura, Karol Bagh, Delhi 110005, India |
We'll acknowledge your grievance within 48 hours and aim to resolve it within 30 days of receiving it. If a complex grievance needs longer, we'll tell you why and when to expect a response, which will be within the maximum period the DPDP Act allows.
Last updated 29 October 2026. These terms are an agreement between you and Dreambots Legal Tech Pvt. Ltd. ("we", "us"), which operates Langdell. They apply when you use langdell.in, join the waitlist or contact us. By using the site, you agree to them.
Please read these terms carefully. If you don't agree with them, please don't use the Site or join the waitlist. If you use the Site on behalf of a firm, court, institution or company, you confirm that you're authorised to accept these terms on its behalf.
You must be at least 18 years old and able to form a binding contract under the Indian Contract Act, 1872 to use the Site or join the waitlist. By using the Site you confirm that you meet these requirements, and that you aren't barred from receiving services under any applicable law.
The Site describes software that we're building and lets you register interest in it. Products, features, workflows, figures and timelines on the Site describe what we're designing and are given in good faith, but they may change, be delayed or be withdrawn before launch. Nothing on the Site is an offer to supply the Service, and nothing on it promises a particular feature, level of accuracy or result.
Illustrations are examples. Product illustrations on the Site, including case titles, party names, deal names, docket numbers, citations, percentages, dates and outcomes, are invented to show how the Service is meant to work. They don't describe real matters or real predictions.
If we invite you to the Beta, your use of it will be governed by a separate beta or customer agreement, which we'll share before you start and which will take precedence over these terms for the Beta. Unless that agreement says otherwise, you should expect that the Beta:
Langdell uses artificial intelligence to read, analyse and summarise legal materials and to produce estimates, drafts and findings. You acknowledge that:
Nothing on the Site is a solicitation or advertisement for legal services, and Dreambots isn't a law firm.
Judgments, orders, statutes, rules, notifications and other public legal materials referred to on the Site or in the Service remain in the public domain, or belong to their respective publishers. Where we add headnotes, summaries, classifications, holdings, citations, links between authorities or other analysis, that added work belongs to us. Our references to a court, tribunal or authority don't imply that it endorses or is affiliated with Langdell.
You agree not to, and not to help anyone else to:
If you discover a security vulnerability, please report it under our responsible disclosure policy rather than testing further.
You're responsible for Your Submissions and confirm that you have the right to send them to us. We use personal data in Your Submissions as described in our privacy policy. You shouldn't send us confidential client information or privileged material through the Site. If you do, we have no obligation to treat it as confidential beyond what the law and our privacy policy require.
If you send us ideas, suggestions or feedback about Langdell, you agree that we may use them freely to improve our products, without paying you and without any obligation to you. We won't publicly identify you as the source without your permission.
The Site and its Content, including the Langdell name, logo and wordmark, the design and the underlying code, belong to Dreambots Legal Tech Pvt. Ltd. or its licensors. They are protected by Indian and international intellectual property law. We grant you a limited, revocable, non-exclusive, non-transferable licence to view the Site and to share links to it for personal or internal professional use. All rights we don't expressly grant are reserved. You may not use our trade marks without our prior written permission, including in domain names, advertising or in a way that suggests we endorse you.
We respect the intellectual property of others. If you believe material on the Site infringes your copyright, send a written notice to our Grievance Officer at info@langdell.in (subject line "Copyright notice"). Include:
We'll review the notice, take appropriate action, including removing or disabling access to the material where justified, and tell you what we've done.
The Site relies on third-party services, such as our host and font provider, and may link to third-party websites such as court portals. We don't control and aren't responsible for third-party websites, their content, availability or privacy practices. A link isn't an endorsement. Your use of a third-party website is at your own risk and subject to its terms.
We aim to keep the Site available, but we don't guarantee that it will be uninterrupted, timely, secure or error-free. We may change, suspend or discontinue all or part of the Site at any time, including for maintenance, without notice or liability.
To the fullest extent the law allows, the Site and its Content are provided "as is" and "as available", without warranties of any kind, express or implied. These include warranties of merchantability, fitness for a particular purpose, accuracy, completeness and non-infringement. We don't warrant that the Site or the Content is accurate, complete, current or free of viruses or other harmful components.
To the fullest extent the law allows:
Nothing in these terms excludes or limits liability for fraud, for death or personal injury caused by negligence, or for any other liability that can't be excluded or limited under Indian law.
You agree to indemnify Dreambots, its directors, employees and agents against any claim, loss, liability, damage or expense (including reasonable legal fees) brought by a third party and arising from your breach of these terms, your misuse of the Site, or Your Submissions.
We may suspend or end your access to the Site or remove you from the waitlist, without notice, if we reasonably believe you've breached these terms, if required by law, or to protect the Site or other users. Sections 6, 7, 9 to 11 and 15 to 21 survive termination.
These terms, and any dispute or claim arising out of or in connection with them or the Site, are governed by the laws of India. Before starting any proceedings, you agree to raise your concern with our Grievance Officer and give us 30 days to try to resolve it. Subject to that, the courts at New Delhi have exclusive jurisdiction.
Complaints about the Site, its Content or these terms can be sent to our Grievance Officer, Avisek Sil, at info@langdell.in, or by post to Dreambots Legal Tech Pvt. Ltd., 4440, 1st Floor, Gali No. 55, Regarpura, Karol Bagh, Delhi 110005. We'll acknowledge complaints within 48 hours and aim to resolve them within 30 days.
Last updated 4 November 2026. In short: we use no analytics, advertising or tracking cookies, which is why you haven't seen a cookie banner.
Cookies are small text files a website asks your browser to store so it can recognise your browser later. Similar technologies, such as local storage, keep information on your device in much the same way. Some are strictly necessary, because a site can't work or stay secure without them. Others are used for analytics, personalisation or advertising, and the law requires a site to ask your permission before using those.
We keep this list short on purpose. This is everything the Site can store on your device:
| Name | Type and provider | Purpose | Duration |
|---|---|---|---|
| langdell-waitlist | Local storage · Langdell | Remembers that you've joined the waitlist and the type of user you chose, so the Site doesn't ask you again. It doesn't hold your email and never leaves your device. | Until you clear your browser data |
| __cf_bm | Cookie · Cloudflare (strictly necessary) | Set only if Cloudflare's security systems detect unusual traffic, to tell people from automated bots. It isn't used to track you or for advertising. | 30 minutes |
The Site follows your device's light or dark setting directly and stores nothing for it.
We don't use analytics cookies, advertising or retargeting cookies, social media pixels, session recording, browser fingerprinting, or any technology that tracks you across other websites. That's why the Site has no cookie banner: nothing it stores on your device needs your consent.
The Site's typefaces are built into the page, so no font provider receives your data. Our host, Cloudflare, handles the technical data needed to deliver every page and measures visits and page speed, as described in our privacy policy. Its measuring script stores nothing on your device.
You can see and delete what the Site has stored at any time:
You can also set your browser to block all cookies. The Site will still work, though if Cloudflare needs to run a security check you may be asked to complete it more often.
If we ever add tools that use cookies or similar technologies for anything beyond running the Site securely, we'll update this policy first. We'll also ask for your consent before those tools load, with "Reject" as easy to choose as "Accept", and let you change your choice at any time.
Questions about this policy: info@langdell.in.
Last updated 4 November 2026. The third-party companies that process personal data on our behalf, what they do, and where.
| Provider | What they do for us | Data they process | Location |
|---|---|---|---|
| Supabase Inc. | Hosts the database that stores waitlist sign-ups and contact messages | Names, email addresses, user type, organisation, product interests, messages | Mumbai, India |
| Cloudflare, Inc. | Hosts and delivers the website, protects it from attacks and abuse, and measures visits and page speed | IP addresses, browser details, request data, page speed measurements | Global network; nearest location to the visitor |
Before we use a provider, we review its security practices, certifications and data protection terms. We engage only providers that:
This list covers the website. Before the private beta opens, we'll publish the subprocessors that will process customer content in the Langdell product, such as infrastructure and AI model providers. Our commitment is that customer content is stored in India and isn't used by any provider to train models that serve others.
We'll update this page before we add or replace a subprocessor. Once the product launches, customers who subscribe to updates will be notified at least 30 days before a new subprocessor begins processing customer content, and may object on reasonable data protection grounds. To ask about a subprocessor, email info@langdell.in.
Langdell's agents read judgments, run diligence and keep matter records. These are the commitments we build them to, and the standard we hold ourselves to before launch.
An answer is only as good as the paragraph behind it. Every holding, estimate and finding links to the exact judgment paragraph, statute section or document page it relied on, so you can check it in seconds.
When Langdell estimates an outcome, it gives a range and the precedents on each side, not a single confident number. When the law is unsettled or the evidence is thin, it says so.
Agents prepare work; lawyers decide. Langdell never sends an email, files a document or writes to your calendar without your explicit approval, and every action is logged so you can see what happened and when.
We won't use your matters, documents or searches to train models that serve other customers. Client confidentiality is the starting point for how we design Langdell.
We test against Indian courts, tribunals, statutes and practice, including the masking of victims' names in sexual-offence and juvenile matters that Indian law requires.
Practising advocates and researchers build our test sets and review where the model goes wrong. We measure citation accuracy and outcome calibration before shipping changes, and we'll publish what we learn.
Langdell supports legal judgment; it doesn't replace it. Advocates remain responsible for the advice they give and the documents they file, and we design the product to make checking easy, not to discourage it.
We take the security of legal data seriously. If you believe you've found a vulnerability in Langdell, we want to hear from you.
Email info@langdell.in with the subject line "Security report". Please include:
Reports from automated scanners without a demonstrated impact, missing best-practice headers with no exploit, and issues in third-party services we don't control. Please report those to the service concerned.
Last updated 29 October 2026. Langdell should be usable by every advocate, judge and student, including people who use assistive technology.
We aim to meet the Web Content Accessibility Guidelines (WCAG) 2.2 at level AA, in line with India's guidelines for accessible digital services.
The interactive background animations on our product pages respond to the mouse and have no keyboard equivalent. They are decorative and carry no information you need. We're also continuing to test with screen readers in Hindi and regional languages.
If something on this site is hard to use, email info@langdell.in with the page and what went wrong. We'll reply within five working days and, where we can, send the information you need in another format.