LexStruct processes any Indian court judgment through 12 pipeline layers — extracting issues, arguments, precedents, applicable law, detecting argument clashes, verifying accuracy via Brave Search, calculating tax exposure, and generating a structured headnote in English and Hindi. Entirely on-premise via Deloitte GenW.AI.
A single Indian court judgment can run 50–200 pages, written in dense legal language with zero structure. Lawyers read the whole document before analysis can begin. That is hours of expensive time — every single case.
Issues, arguments, precedents and verdict are buried in dense paragraphs — invisible unless you read every word.
Lawyers spend hours getting through the document before any analysis begins. Expensive and unsustainable at scale.
Without systematic search, similar past cases that strengthen or weaken an argument go entirely unnoticed.
Insights from past cases stay locked in lawyers' heads, not in a searchable, growing database everyone can access.
GenW.AI is Deloitte's enterprise agent platform. LexStruct deploys multiple specialised agents across four GenW modules — all on-premise inside your organisation's network. No data leaves your infrastructure.
A multi-agent pipeline built on Deloitte GenW.AI — four modules working together to read, understand, and structure any Indian court judgment in under 2 minutes. Each agent has one specific job. All processing stays on-premise inside your organisation's network.
Every agent is built on Deloitte GenW.AI's Agent Builder module and runs inside the LexStruct pipeline. Here is what each one actually does and why it matters.
The full multi-agent LexStruct pipeline. Each agent has one job. All run on-premise inside the Deloitte GenW.AI network — no case data leaves your organisation.
| # | Agent Name | GenW Module | LLM / Tool | Role | Tag |
|---|---|---|---|---|---|
| 01 | Realm AI Ingestion | Realm AI | 2K-token chunker | Chunks PDF, builds Realm AI knowledge base | |
| 02 | Master Orchestrator | Agent Builder | Parallel router | Routes chunks to 5 specialist agents in parallel, manages flow state | |
| 03 | Issue Agent | Agent Builder | Sarvam / GPT-4 | Extracts core legal question from judgment | |
| 04 | Petitioner Agent | Agent Builder | Sarvam / GPT-4 | Extracts petitioner arguments exclusively, isolated from respondent context | I7 |
| 05 | Respondent Agent | Agent Builder | Sarvam / GPT-4 | Extracts respondent arguments exclusively, runs parallel to Agent 04 | I7 |
| 06 | Precedent Agent | Agent Builder | LexGraph N1 | Extracts all cited cases and applicable law sections | |
| 07 | Judgment Agent | Agent Builder | Sarvam / GPT-4 | Extracts court reasoning chain and preliminary verdict | |
| 08 | Argument Clash Detector | Agent Builder | GPT-4 comparison | Labels P vs R: Contradicted · Conceded · Novel | N3 |
| 09 | Good Law Checker | Agent Builder | LexGraph Graph traversal | Walks LexGraph, flags overruled / distinguished cases | N9 |
| 10 | VL Query Builder | Agent Builder | GPT-4 | Converts extracted field to Brave Search query | I3 |
| 11 | Compare LLM Agent | Agent Builder | GPT-4 Brave Search API | Scores agent output vs web content 0–100% | I1 I3 |
| 12 | HITL Condition Node | Agent Builder | Condition node | Fires HITL review only when any field score < 80% | I2 |
| 13 | Synthesis + Verdict | Agent Builder | Sarvam / GPT-4 | Assembles headnote, tags verdict Allowed/Dismissed/Partial | I8 |
| 14 | Citation Strength Scorer | Agent Builder | LexGraph PageRank | PageRank-style authority score per cited case | N2 |
| 15 | Counter-Arg Generator | Agent Builder | RAG · PostgreSQL | RAGs DB for top 3 adversarial responses per argument | N5 |
| 16 | Vernacular Agent | Agent Builder | IndicTrans / Sarvam | Translates headnote to Hindi / regional language | N10 |
| 17 | SQL Generation Agent | Agent Builder | SQL gen PostgreSQL | Engineers INSERT queries, stores all fields with confidence scores | I6 |
| 18 | Tax Demand Calculator | Agent Builder | PostgreSQL | Computes rupee exposure range from similar case outcome patterns | N7 |
| 19 | Section Lifecycle Agent | Agent Builder | PostgreSQL · LexGraph | Year-on-year interpretive drift per law section | N4 |
Every screen your lawyers, analysts, and managers will use — built inside Deloitte GenW App Maker. The platform is named LexStruct in the product UI and connects directly to the full multi-agent pipeline. No technical knowledge needed to use any of it.
LexStruct Pipeline Ready — Multi-Agent