VerdictIQIndian Legal AI Workbench

Research Indian Law
With AI.

Understand case facts. Extract legal issues. Find relevant Supreme Court and High Court precedents on Indian Kanoon instantly.

Built in collaboration with practicing Indian Advocates

Designed for Indian Legal Practice

Litigation ChambersIndependent AdvocatesAppellate CounselTrial Court Advocates

How Advocates Research Today

Legal research in India is repetitive and time-consuming. Because traditional search engines look for literal keywords, advocates spend hours reading irrelevant judgments.

01

Read Client Facts

Manually digest case papers to identify dockets and facts.

02

Search Indian Kanoon

Enter keywords, filtering through hundreds of identical search hits.

03

Open 40 Rulings

Read dozens of judgments across multiple browser tabs to compare fact details.

04

Lose Essential Hours

Spend days isolating court holdings and ratios before drafting briefs.

How it Works

AI-Powered Research Workflow

From client facts to matching precedents in six steps. Simple, intuitive, and secure.

01

Paste Case Facts

Paste client details or raw text. VerdictIQ parses the context immediately.

02

Extract Parameters

The model extracts case types, core issues, central Acts, sections, and keywords.

03

Kanoon Search & Re-ranking

Queries are sent to Indian Kanoon. Results are re-ranked by LLM fact similarity.

AI Case Understanding

Realistic Parameter Mapping

See how our model processes a RERA delayed flat possession dispute in Phase 1.

Advocate Fact Input
"Builder delayed possession of flat by four years and refused refund."
Mapped Acts & Sections
Consumer Protection ActRERA Act, 2016Section 18 RERA
Extracted Issues & Keywords
Builder DefaultDelay in PossessionRefund ClaimsApartment Delivery

Supported Pleadings

Phase 1 supports text parsing for standard court documents. Scanned PDFs and OCR processing belong to Phase 2.

PetitionsActive
ComplaintsActive
Legal NoticesActive
Court OrdersActive
Written StatementsActive
FIR DocumentsActive
Scanned PDFsPhase 2
DOCX / ImagesPhase 2

Search Coverage

Research is conducted across the primary Indian judicial dockets indexed on Indian Kanoon.

Supreme Court of IndiaFully Covered (Phase 1)
All State High CourtsFully Covered (Phase 1)
National Tribunals (NCLT, NGT, etc.)Phase 2
District & Consumer CourtsPhase 5
Semantic Search

Designed for Indian Law

How VerdictIQ semantic mapping compares to legacy portal keyword searches.

Legacy Keywords
  • Rigid keyword requirements that miss synonyms.
  • Manual filtering of dockets by date and hits only.
  • Requires reading entire PDF files to find holdings.
VerdictIQ Semantic AI
  • AI Case Fact understanding that maps semantic meaning.
  • Re-ranks search results based on fact similarity.
  • Dynamic AI summaries isolating ratio decidendi.

Built for Every Chambers

Leveling the research field for independent bars and junior advocates.

Junior Lawyers

Build confidence quickly. Isolate court holdings, ratio decidendi, and precedent arguments in seconds to brief seniors effectively.

  • Ratio decidendi extraction
  • Static docket summaries
  • Act parameter mapping
Chambers Favorite

Litigation Chambers

Maximize search speed. Coordinate research folders and search histories across senior partners and junior associates.

  • Shared bookmark folders
  • Unlimited search history
  • Joint research files

Legal Researchers

Conduct deep jurisprudential analysis. Trace citations and identify statutory splits across High Courts quickly.

  • Judicial citation mapping
  • Multi-bench search comparison
  • Statute hierarchy tracking

Honest about your data

VerdictIQ is an academic project. Facts you submit are sent to a hosted model API to generate answers, and your research is saved to your account so you can return to it. Use anonymised facts, and delete sessions whenever you like.

You control your history

Delete any research session, or clear them all, from History in the app.

Secure sign-in

Passwords stored only as bcrypt hashes; refresh token in an HttpOnly cookie.

No training on your data

Nothing you submit is used to train or fine-tune any model.

Frequently Asked Questions

What is AI legal research?
AI legal research uses natural language understanding (NLP) to analyze the factual context of your case rather than looking for literal keyword matches. It helps you identify legal issues, statutes, and relevant precedents that keyword searches miss.
How is VerdictIQ different from Indian Kanoon?
Indian Kanoon is an excellent judgment index. VerdictIQ acts as an intelligent layer on top of it. Instead of typing boolean operators and keyword filters, you input your case facts in plain English. The AI parses the facts, maps relevant central or state Acts/sections, queries the index, and re-ranks the results based on actual fact similarity.
Does the AI hallucinate fake cases?
No. Unlike generic conversational LLMs, VerdictIQ employs strict anti-hallucination protocols. Every case cited in our dashboard is grounded in official, verified dockets and links directly back to its source text on Indian Kanoon.
What courts and jurisdictions are covered in Phase 1?
Phase 1 covers all judgments from the Supreme Court of India and all State High Courts indexed by the Indian Kanoon API.
Can I upload files like petitions or FIRs?
Yes. In Phase 1, you can copy and paste raw facts directly or upload text-based PDFs (such as petitions, written statements, FIRs, court orders, or legal notices). Scanned PDFs and image files requiring OCR will be supported in Phase 2.
Will my client data remain private?
Yes. We run a zero-retention infrastructure. Pleadings and fact sheets uploaded to our servers are cached temporarily for parameter mapping and similarity calculations, and are deleted immediately after the session. We never store them permanently or use them to train shared models.
What is a similarity score?
A similarity score (expressed as a percentage) indicates how closely the facts and legal reasoning of a retrieved precedent match the facts of your active case, helping you prioritize what to read first.
How are judgments ranked?
Your facts are searched against our own index of Supreme Court judgments (meaning-based and keyword search combined), with live Indian Kanoon search as a supplement. Each precedent then gets a measured 5-dimension score — factual, statutory, procedural, relief and ratio similarity — and the most similar appear first. The app shows how every score was computed.
What is Section 18 of RERA?
Section 18 of the Real Estate (Regulation and Development) Act, 2016, outlines the homebuyer's right to claim a refund and interest from a developer in case of delayed possession of a property.
Which AI models does VerdictIQ use?
Open-weight language models (such as gpt-oss and Qwen) served through the Groq API, with other OpenAI-compatible providers as fallbacks. Because these are hosted APIs, the facts you submit are sent to the provider — use anonymised facts. Search embeddings (BAAI/bge-small-en-v1.5) run on our own server. If no model is available, VerdictIQ tells you so instead of showing a made-up answer.
Can I use this for Section 138 Negotiable Instruments Act cases?
Yes. Our models are optimized to extract check bounce parameters, such as legally enforceable debt presumptions, statutory notice delivery, and timeline defaults, linking you directly to SC precedents like Rangappa.
Is this tool a replacement for lawyers?
Absolutely not. VerdictIQ is an assistant designed to automate the repetitive hours of precedent searching. It provides research summaries, but final statutory analysis, citation verification, and court representation remain the sole responsibility of the advocate.
Can I use VerdictIQ for state-specific laws?
Yes, our model extracts references to state-specific amendments and local rules (e.g. state RERA rules, local municipality Acts) and incorporates them into search queries.
Is there a limit to how many searches I can perform?
Research capacity is managed via account access tiers. Free access includes monthly evaluation credits, while higher tiers support extended research workloads.
How do I download research reports?
Reports containing your uploaded facts, extracted legal issues, matching precedents, and AI summaries can be exported as clean PDFs directly from the dashboard.
What is the project access model?
VerdictIQ provides tiered research access. Users can create an account and immediately explore precedent analysis in the interactive workbench.
Can my junior associates access my search history?
Under the collaborative tier, chambers can create shared workspaces allowing junior advocates and senior partners to coordinate search histories and saved cases.
Can I search for district court judgments?
VerdictIQ currently indexing covers the Supreme Court and High Courts of India.
How quickly can I access the platform?
Access is instant. Once you register your account and choose a plan, you can immediately begin inputting case facts and extracting precedents.
How do I get in touch with the team?
You can email Team Lead Partha Shankar (parthashankaer21@gmail.com) directly or use our contact page.

Ready to research
faster with AI?

VerdictIQ is fully launched. Sign in or register now to start analyzing case facts and retrieving precedents.