5-Year Integrated BBA LLB
Program Design Blueprint

Semester I · Compulsory Law Paper · 4 Credits · 3-0-2-0 · Moot-blended. Statutory basis: Indian Contract Act 1872; Specific Relief Act 1963 (overview); IT Act 2000 (e-contracts). Aligned to BCI Rules of Legal Education 2008; benchmarked against NLSIU, NALSAR, NLU Delhi and JGLS.

1 · Course Overview

The Law of Contracts is the doctrinal and commercial bedrock of a lawyer's training. It governs the legally enforceable promises that structure every transaction — from a simple sale to a complex cross-border financing. The course develops a rigorous understanding of the Indian Contract Act, 1872, from the anatomy of agreement (offer, acceptance, consideration, capacity, free consent, legality) through performance, discharge, breach and remedies. Because this is a B.B.A. LL.B. cohort, it is taught with a deliberate commercial lens: students learn not only what the law is, but how contractual risk is identified, allocated and drafted around in real transactions.

The course is consciously skills-forward. Doctrine is delivered alongside drafting, clause analysis, negotiation and oral advocacy, so that a student leaving Semester I can read a contract critically, spot a defective term, draft a clean clause, and argue a contractual dispute on its facts — the difference between a knowledgeable graduate and a placement-ready one.

Course Objectives

1. Establish a precise, statute-anchored understanding of formation, validity, performance and discharge of contracts.
2. Develop analytical ability to apply contractual doctrine to novel commercial fact-patterns and predict judicial outcomes.
3. Build foundational drafting and clause-analysis skills that translate doctrine into enforceable, risk-allocated text.
4. Cultivate oral advocacy and negotiation through structured moot and simulation exercises grounded in contract disputes.
5. Introduce the technology and tools modern transactional practice uses to research, draft and review contracts.

2 · Course Outcomes, BTL & PO/PSO Mapping

BTL key: L2 Understand · L3 Apply · L4 Analyse · L5 Evaluate (Bloom's revised taxonomy).

CO → PO / PSO Correlation (3 High · 2 Medium · 1 Low)

PSO1 = Commercial & transactional acumen (the BBA-LLB business-law blend). PSO2 = Legal-technology & drafting fluency.

Rationale: CO5 and CO6 carry the highest PSO weights because drafting (PSO2) and commercial risk-allocation (PSO1) are exercised most directly there. CO6 maps High to PO8 (case management & teamwork) and PO9 (communication) as the moot is a team-based oral-advocacy outcome. PO10 (lifelong learning) rises across CO4–CO6 as students engage evolving case law and e-contract practice.

3 · Detailed Syllabus — Six Modules

Six modules, each mapped consecutively to one Course Outcome (Module 1 → CO1 … Module 6 → CO6). The flow proceeds logically from formation, through validity and special categories, to performance, breach, remedies, drafting and finally dispute advocacy.

4 · Case Studies & Landmark Cases (Module-Mapped)

Case Study A — The Startup Co-Founder Note Module 2

Student tasks

1. Identify each promise and test whether valid consideration supports it (past/present/future; adequacy vs. sufficiency).
2. Assess enforceability of the minor's promise applying Mohori Bibee and S.10–S.11; advise on the necessaries exception.
3. Apply privity to the mentor's claimed benefit; identify any applicable exception.
4. Redraft the founder note into a legally sound short agreement curing the defects.

Deliverables

• A 600–800 word legal opinion advising the eldest founder on enforceability.
• A redrafted, defect-free 'Founders' Agreement' (1–2 pages) with a note explaining each correction.

Learning outcomes

• Apply consideration and capacity doctrine to a live commercial dispute (CO2).
• Experience the doctrine-to-drafting bridge anchoring PSO1/PSO2.
• Recognise how privity operates in multi-party startup arrangements.

Case Study C — The Failed Supply Contract (COVID Force-Majeure) Modules 4 & 6

Student tasks

1. Determine whether the contract is frustrated under S.56 or merely more onerous (Satyabrata Ghose; Energy Watchdog v. CERC).
2. Construe the force-majeure clause: does 'government action' cover the lockdown, and does it excuse or merely suspend performance?
3. Quantify the distributor's damages (Hadley v. Baxendale) and the duty to mitigate; assess any liquidated-damages clause.
4. Draft the legal notice the distributor would issue, and outline the manufacturer's defence (re-used in the Module 6 moot).

Deliverables

• A two-sided issue memo (manufacturer vs. distributor), ~1,000 words.
• A drafted legal/demand notice from the distributor.
• A one-page damages computation sheet with mitigation adjustments.

Learning outcomes

• Evaluate competing remedies and frustration arguments on real facts (CO4).
• Bridge breach doctrine into written advocacy and quantification (CO5/CO6).
• Engage current force-majeure jurisprudence (PO10).

Case Study D — The SaaS Master Services Agreement Redline Module 5

Student tasks

1. Read the extract adversarially from the customer's side; identify and rank the five most dangerous clauses.
2. Redline each problematic clause with a one-line negotiation comment justifying each change.
3. Apply doctrine from Modules 3–4 (S.23, S.27, limitation of liability, penalty vs. liquidated damages) to support the redlines.
4. Produce a clean, balanced version of the six clauses.

Deliverables

• A redlined version with margin comments (tracked changes).
• A 'clause-risk register': clause | risk | doctrine engaged | proposed fix.
• A clean redrafted clause set.

Learning outcomes

• Construct and critique clauses and allocate risk (CO5; PSO2 High).
• Operate the adversarial drafting mindset placements test directly.
• Use a clause/CLM tool to perform and track the redline.

Case Study B — The Coerced Settlement Module 3

Student tasks

1. Identify the vitiating factors potentially in play (coercion, undue influence, economic duress) and their statutory basis.
2. Apply the law to the facts: was the threat legitimate, and was consent truly free?
3. Distinguish hard bargaining from unlawful pressure using authority.
4. Advise whether the settlement is voidable and what remedy follows.

Deliverables

• An issue analysis of the vitiating factors.
• An application of authority to the facts.
• A reasoned advice on voidability and remedy.

Learning outcomes

• Analyse free consent and its vitiating factors (CO mapping: formation/validity).
• Distinguish legitimate commercial pressure from unlawful coercion.
• Advise on the consequences of vitiated consent.

Case Study E — The Distributor's Anticipatory Breach Module 6

Student tasks

1. Determine whether the manufacturer's letter amounts to anticipatory breach.
2. Analyse the innocent party's election: accept the repudiation and sue immediately, or affirm and wait.
3. Assess the measure of damages, including mitigation and remoteness.
4. Argue the case for each side before the moot bench on written and oral submissions.

Deliverables

• A case analysis identifying anticipatory breach and the election.
• Written submissions for one side.
• Oral advocacy before the moot bench (assessed under the moot rubric).

Learning outcomes

• Analyse breach, repudiation and the innocent party's options (CO mapping: breach/remedies).
• Apply mitigation and remoteness to quantify damages.
• Advocate a contract position under examination in a moot setting.

5 · Active Learning Methods

Four high-impact, course-relevant ALMs are embedded; trivial methods are excluded. Each has an implementation protocol and a course-specific rubric.

ALM 1 · Socratic Case Dialogue

A structured cold-call dialogue interrogating a landmark case (Carlill, Mohori Bibee, Hadley) through progressively harder hypotheticals, forcing students to defend, refine and abandon positions in real time — the core method of elite law schools. Implementation: pre-read case briefs → 20–25 min escalating dialogue per anchor case → rotating scribe maps the reasoning chain → 3-minute written exit reasoning note.

ALM 2 · Clause Drafting & Adversarial Redline Studio

Students draft a clause to a brief, then exchange and redline a peer's clause from the opposing party's perspective — the single most placement-relevant skill. Implementation: drafting brief issued → 20-min draft with defined-terms discipline → swap and redline from opposing perspective → original drafter responds (accept/reject with reasons) → instructor debriefs the best fix.

ALM 3 · Moot / Oral Advocacy Simulation

Structured oral advocacy on a contract dispute before a bench, with memorial exchange — builds courtroom presence, argument structure and thinking on one's feet. Delivered through the Section 9 moot exercises. Implementation: side assigned → short memorial filed → 10-min oral rounds + 3-min rebuttal before a questioning bench → immediate oral feedback + written score sheet.

ALM 4 · Problem-Based Learning — Transaction Simulation

Teams work a multi-week realistic scenario (the Section 4 case studies) end-to-end — diagnosing issues, drafting documents, advising a 'client' — integrating doctrine across modules. Implementation: scenario released at module start → staged deliverables (issue memo → drafted document → client advice) with weekly checkpoints → mid-point peer review → 10-min client clinic presentation.

6 · Course-Based Team Activities

Specific to Law of Contracts and distinct from the semester-wide Activity Stream course. Teams of 4–5; every activity builds a core lawyering ability (draft, summarise, argue, present, debate, question), with drafting and pleading emphasised for placements.

7 · Reference Books & MOOCs

Prescribed Textbooks

Reference Books

Recommended MOOCs

8 · Global Certifications & National Benchmarking

Relevant Certifications (none forced)

Heavier certifications (CS-ICSI, CIArb, IAPP) attach to later specialised courses — not to this foundational paper.

Benchmarking Against Nationally Reputed Universities

Reasoning for missed topics: the ~8–12% not covered is a deliberate architectural choice, not a gap — indemnity, guarantee, bailment, pledge and agency move to Special Contracts (Sem II); Specific Relief is introduced here only as a remedy and treated fully later; deep economic analysis is met through the BBA economics papers. This avoids duplication and respects the integrated programme.

9 · Moot Court / Oral-Advocacy Exercises (Moot-Blended)

As a moot-blended course, 2 contact hours/week are dedicated to structured oral-advocacy and drafting exercises. Ten exercises run across the semester, escalating from foundational skills to a full moot. Each releases a problem ≥48 hours ahead; written work precedes the oral round; a questioning bench returns a standard score sheet; exercises are cumulative; the final two are graded and feed the moot-court assessment component.

Moot Evaluation Rubric (per exercise)

10 · Technology Blend & Tools

Technology is blended lightly and purposefully — only where it sharpens a contract-law skill. Two tools are used; no more. This is a doctrinal-plus-drafting course, not a technology course, so it has no lab-experiment component (the 10-lab requirement applies only to Tech-Stream courses).

• Legal research on a live database — students find, validate and cite contract authorities the way practice demands.
• Clause analysis & redlining on a contract/CLM tool — version control, clause libraries and AI-assisted review, the actual associate workflow.
• Where AI assistance is used, students are trained to verify output against statute and case law — building critical verification, not dependence.

Semester II · Technology Stream · 3 Credits · 2-0-2-0 · Technology-blended (lab-based). No statutory basis — a skills-and-tools course. Aligned to the BCI emphasis on practical training and to NITI Aayog / BCI signals on AI in legal practice. Benchmarked against NLSIU, NLUD, Jindal (legal-tech electives) and global legal-tech curricula.

1 · Course Overview

AI-Augmented Legal Research teaches the modern research workflow that has replaced passive reading in professional practice. A first-year law student must learn to find the right authority fast, validate that it is still good law, read it critically, and increasingly to direct and verify AI research assistants rather than be replaced by them. This course builds that competence on live legal databases and current legal-AI tools, in a hands-on lab format.

The course is explicitly anti-hallucination: every AI output a student generates must be verified against a primary source. Students learn that the value a lawyer adds in an AI era is judgment, verification and the framing of the right question — not raw retrieval. By the end, a student can run a full research trail on a real legal problem, produce a verified research note, and explain exactly how an AI assistant helped and where it could not be trusted.

Course Objectives

1. Build fluency in primary and secondary legal-research sources and the Indian citation system.
2. Develop a disciplined, reproducible research methodology — from issue to authority to verified note.
3. Teach effective, critical use of AI legal-research assistants with mandatory source verification.
4. Establish the habit of checking whether an authority is still good law (noting-up / citator discipline).
5. Produce professional research work-product (research notes, case digests, authority tables) to a practice standard.

2 · Course Outcomes, BTL & PO/PSO Mapping

BTL key: L2 Understand · L3 Apply · L4 Analyse · L5 Evaluate.

CO → PO / PSO Correlation (3 High · 2 Medium · 1 Low)

PSO1 = Commercial & transactional acumen. PSO2 = Legal-technology & drafting fluency (this course is a primary driver of PSO2).

Rationale: CO5 maps High to PO6 (ethics) because judging when AI may be relied upon is fundamentally an ethical-professional decision. PSO2 is High across almost every CO because this is the foundational legal-technology course. PO5 (legal research & reform) is High throughout — it is the course's spine. PO9/PO6 rise at CO6 as work-product is communicated and certified as reliable.

3 · Detailed Syllabus — Six Modules

Six modules, each mapped consecutively to one Course Outcome (Module 1 → CO1 … Module 6 → CO6). The flow runs from sources and citation, through database research methodology and validation, into AI-assisted research and its critical evaluation, ending in professional work-product. Each module is paired with the lab experiments in Section 9.

4 · Case Studies (Module-Mapped)

Case Study A — The Tenant's Question Modules 1 & 2

Student tasks

1. Convert the question into legal issues and a research plan (issue tree, search terms).
2. Identify the relevant sources: the applicable Rent/Tenancy statute, the Contract Act provisions, and case law on forfeiture of deposit.
3. Run the search on SCC Online/Manupatra; capture the research trail.
4. Produce a short sourced answer with citations.

Deliverables

• A one-page research plan (issues + search terms).
• A documented research trail (queries → results → decisions).
• A 400-word sourced answer to the client's question.

Learning outcomes

• Apply structured research methodology to a real query (CO2).
• Map a lay question onto primary sources (CO1).
• Produce a reproducible research trail (CO6 foundation).

Case Study C — The Hallucinated Citation Modules 4 & 5

Student tasks

1. Verify every one of the six citations against a primary source; classify each as accurate, misattributed, or fabricated.
2. Document exactly how each error would have harmed a client or misled a court.
3. Rewrite the note's conclusion based only on verified, correctly-represented authority.
4. Draft a one-paragraph 'AI-use disclosure' explaining how the tool was used and verified.

Deliverables

• A verification table: citation | claimed holding | actual status | verdict.
• A corrected conclusion grounded only in verified authority.
• A short reflection on the professional-responsibility risk.

Learning outcomes

• Operate the mandatory verification loop (CO4).
• Evaluate AI reliability and the ethics of reliance (CO5; PO6 High).
• Internalise why verification is non-negotiable — the course's central lesson.

Case Study B — The Overruled Precedent Trap Module 3

Student tasks

1. Note up the cited authority and trace its full subsequent treatment.
2. Determine its current status (good law, doubted, read down, or overruled) with evidence.
3. Identify the controlling current authority that should replace or qualify it.
4. Rewrite the relevant passage so it relies only on currently good law.

Deliverables

• A noting-up trail for the authority.
• A validity verdict with supporting citations.
• A corrected passage relying on current good law.

Learning outcomes

• Analyse an authority for current validity (CO3).
• Internalise why citing overruled law is a career-defining error.
• Document validity so a reader can trust the work.

Case Study D — The 24-Hour Research Memo Module 6

Student tasks

1. Frame the issue and build a rapid research plan.
2. Research and validate the controlling authorities, using AI assistance where it genuinely accelerates.
3. Verify every AI-surfaced citation against a primary source.
4. Write an answer-first memo with a clear conclusion, reasoning, authorities and an AI-use disclosure.

Deliverables

• A research trail.
• A verified list of authorities.
• A professional, answer-first research memo with an AI-use disclosure.

Learning outcomes

• Construct professional research work-product under pressure (CO6).
• Integrate safe AI use with verification (CO4).
• Communicate uncertainty honestly where the law is unsettled.

5 · Active Learning Methods

Four high-impact, course-relevant ALMs. Trivial methods are excluded in favour of methods that build genuine research-and-verification competence.

ALM 1 · Live Research Race

A timed, competitive research challenge: given a legal question, teams race to find the controlling authority and confirm it is good law, presenting their trail. Builds speed, search instinct and the validation reflex. Implementation: question released live → 25-minute timed search on the database → teams present the controlling authority + proof of currency → fastest correct-and-verified trail wins.

ALM 2 · AI-vs-Human Verification Duel

Students run a research question through an AI assistant, then independently verify it on the database, scoring where the AI helped and where it failed — building the critical-verification habit that defines safe AI use. Implementation: same question to AI and to manual research → side-by-side comparison table → class debrief on every divergence → tally of AI hits, misses and hallucinations.

ALM 3 · Research-Trail Peer Review

Students exchange research notes and reproduce each other's trail to test whether the conclusion holds — the legal equivalent of code review, building rigour and reproducibility. Implementation: submit note + trail → partner attempts to reproduce the result → partner files a reproduction report (confirmed / could-not-reproduce / found-better-authority) → original author responds.

ALM 4 · Problem-Based Research Lab

Teams take a multi-part legal problem and build a complete, AI-assisted-but-human-verified research dossier over the lab sequence — integrating every skill in the course. Implementation: problem released → staged lab deliverables (plan → authorities → validation → AI-assist → verified note) → checkpoints in each lab → final dossier presented.

6 · Course-Based Team Activities

Specific to this course and distinct from the semester-wide Activity Stream. Teams of 4–5; each builds a lawyering ability (research, summarise, present, question, certify) with research-and-verification rigour emphasised.

7 · Resources & MOOCs

Prescribed / Reference Texts

Recommended MOOCs

8 · Global Certifications & National Benchmarking

Relevant Certifications (none forced)

As a foundational tech course, heavyweight certs are deferred; vendor research badges are the natural, low-cost fit here.

Benchmarking Against Nationally Reputed Universities

Reasoning for missed topics: empirical/quantitative research methods move to the Sem-IV Legal Research Methodology lab; document automation and data science move to TECH-202; this course deliberately concentrates on the find-validate-verify research spine so it is mastered before later courses build on it.

9 · Laboratory Experiments (Technology Course)

As a technology-blended course, TECH-102 is delivered through a structured lab. Ten experiments run across the semester in the 2-hour weekly lab slot, each producing an artifact. Conduct: each lab has a pre-lab brief, an in-lab task on a live tool, and a submitted artifact; labs are cumulative and feed the Module-6 capstone dossier; a standard lab rubric applies (below).

Lab Evaluation Rubric (per experiment)

10 · Technology Blend & Tools

This is a technology course, so technology is the medium, not an add-on. Two primary tool families are used; the course resists tool sprawl so students go deep, not wide.

• Legal-research databases are the core environment — every lab runs on a live database so retrieval, filtering and noting-up become muscle memory.
• AI research assistants are introduced as accelerators under a strict verification regime — students learn to direct and check them, building the human-in-the-loop discipline.
• The course explicitly teaches the limits of the technology — bias, hallucination, privacy — so fluency is paired with judgment.

Semester I · Technology Stream · 2 Credits · 2-0-2-0 · Technology-blended (lab-based). The foundation course of the Technology Stream — builds the computational mental model every later tech course relies on. No statutory basis. Benchmarked against NLSIU, NLUD, JGLS legal-tech foundations and global "law + computation" curricula.

1 · Course Overview

Legal Informatics & Computational Thinking is the entry point to the Technology Stream. Before a law student can use AI tools, govern algorithms or build legal-tech, they need a working mental model of how computers represent information, how data is structured, and how legal rules can be expressed as logic. This course builds that model from zero, assuming no prior coding background, using law-native examples throughout.

The course is deliberately conceptual-plus-hands-on rather than a programming bootcamp. Students learn computational thinking — decomposition, pattern recognition, abstraction and algorithm design — and apply it to legal problems: modelling a statute as a decision tree, representing a contract as structured data, and understanding how legal databases and APIs actually work. It demystifies technology so the rest of the stream is approachable.

Course Objectives

1. Develop computational-thinking skills (decomposition, abstraction, pattern recognition, algorithms) applied to legal problems.
2. Build a working understanding of data, data structures and how legal information is digitally represented.
3. Introduce 'rules as code' — expressing legal rules as machine-readable logic.
4. Familiarise students with legal ontologies, APIs and the architecture of legal-tech systems.
5. Demystify technology so students approach the rest of the Technology Stream with confidence.

2 · Course Outcomes, BTL & PO/PSO Mapping

BTL key: L2 Understand · L3 Apply · L4 Analyse · L5 Evaluate.

CO → PO / PSO Correlation (3 High · 2 Medium · 1 Low)

PSO1 = Commercial & transactional acumen. PSO2 = Legal-technology & drafting fluency.

Rationale: PSO2 (legal-technology fluency) is High across every CO — this is the stream's foundation course. PO5 (legal research & reform) rises where students model rules as code. CO5 maps to PO6 (ethics) because judging where automation is appropriate is a professional-responsibility question.

3 · Detailed Syllabus — Six Modules

Six modules, each mapped consecutively to one Course Outcome (Module 1 → CO1 … Module 6 → CO6). Each module is paired with the lab experiments in Section 9.

4 · Case Studies (Module-Mapped)

Case Study B — Encoding the Limitation Period Module 3

Student tasks

1. Decompose the provision into its conditions, exceptions and date computations.
2. Draw a decision tree capturing every path, including acknowledgement and disability exclusions.
3. Encode it as a simple rule model (no-code or pseudo-code) and test it on five fact-patterns.
4. Document the edge cases where the model is uncertain and a lawyer must intervene.

Deliverables

• A decision-tree diagram of the limitation rule.
• A rule model that outputs time-barred / not time-barred.
• A note on the edge cases requiring human judgment.

Learning outcomes

• Analyse a statute and represent it as structured logic (CO3).
• Experience why exceptions make legal codification hard.
• Build the rules-as-code foundation for compliance automation.

Case Study C — The Messy Cause-List Modules 4 & 5

Student tasks

1. Identify the entities hidden in the unstructured text (case no., parties, date, court, stage).
2. Design a structured schema (table/JSON) to hold them.
3. Clean and structure a sample of the data into that schema.
4. Reflect on what a downstream tool could and could not do with the cleaned data.

Deliverables

• A proposed data schema.
• A cleaned, structured sample dataset.
• A short reflection on data-quality lessons.

Learning outcomes

• Apply structured-data and ontology thinking (CO4).
• Evaluate the limits of computation on messy legal data (CO5).
• Understand why 'garbage in, garbage out' governs legal-tech.

Case Study A — The Eligibility Maze Modules 1 & 2

Student tasks

1. Decompose the scheme into its atomic conditions, categories and exceptions.
2. Identify overlaps and contradictions in the rules as written.
3. Express the eligibility logic as a structured flow (decision points and branches).
4. Test the logic against five sample applicants and confirm each gets the correct result.

Deliverables

• A decomposition of the scheme into atomic rules.
• A structured eligibility flow (decision points).
• A test sheet of five applicants with expected vs. produced outcomes.

Learning outcomes

• Apply decomposition and abstraction to a messy legal rule-set (CO2).
• Recognise patterns and contradictions across conditions (CO1).
• Build the disciplined logic that any later rules-as-code work depends on.

Case Study D — The Decision-Tree App Module 6

Student tasks

1. Select a provision and write a one-line specification of what the tool must decide.
2. Encode the rule as a working decision-tree / no-code model.
3. Test it against real fact-patterns, including deliberate edge cases.
4. Document every assumption and the precise points where a human lawyer must still intervene.

Deliverables

• A working decision-tree tool.
• A test log including edge cases.
• A limits-and-assumptions note identifying where human judgment is required.

Learning outcomes

• Construct a computational legal artifact end-to-end (CO6).
• Evaluate where automation is and is not appropriate (CO5).
• Communicate a build and its limits to a non-technical audience.

5 · Active Learning Methods

Four high-impact, course-relevant ALMs. Trivial methods are excluded in favour of methods that build genuine technical-and-judgment competence.

ALM 1 · Unplugged Algorithm Design

Students design step-by-step 'algorithms' for legal procedures on paper/whiteboard before any computer is involved — building computational thinking without the friction of code. Implementation: a legal procedure is posed (e.g., 'decide if a person qualifies for legal aid') → teams write the precise step-by-step logic → another team 'executes' it literally to expose gaps → debrief on ambiguity and missing cases.

ALM 2 · Rules-as-Code Studio

Hands-on studio where students encode a legal rule as a decision tree / no-code model and test it against fact-patterns — the core technical skill of the course. Implementation: a provision is assigned → students build the model in a no-code/diagram tool → swap and test a peer's model with adversarial fact-patterns → fix and document.

ALM 3 · Tech-Claim Teardown

Students critically dissect a real legal-tech product's marketing claims against what is technically plausible — building the evaluator's eye that protects clients and firms from hype. Implementation: a product page/claim is given → teams map each claim to the underlying capability → classify as real / exaggerated / impossible → present the verdict.

ALM 4 · Build-and-Demo (PBL)

Teams build the Module-6 capstone artifact over the lab sequence and demo it to a non-technical 'client' — integrating every skill end-to-end. Implementation: pick a provision → staged build (spec → logic → test → document) with lab checkpoints → live demo + Q&A from a non-technical audience.

6 · Course-Based Team Activities

Specific to this course and distinct from the semester-wide Activity Stream. Teams of 4–5; each builds a lawyering ability with technical-and-verification rigour emphasised.

7 · Resources & MOOCs

Prescribed / Reference Texts

Recommended MOOCs

8 · Global Certifications & National Benchmarking

Relevant Certifications (none forced)

Benchmarking Against Nationally Reputed Universities

Reasoning for missed topics: The ~12–20% not covered is intentional: actual programming, data science and ML move to TECH-202; this course deliberately stays no-code and concept-first so every law student — regardless of background — gains the computational mental model before any coding is required.

9 · Laboratory Experiments (Technology Course)

As a technology-blended course, this course is delivered through a structured lab. Ten experiments run across the semester in the 2-hour weekly lab slot, each producing an artifact. Conduct: each lab has a pre-lab brief, an in-lab task on a live tool, and a submitted artifact; labs are cumulative and feed the Module-6 capstone; a standard lab rubric applies (below).

Lab Evaluation Rubric (per experiment)

10 · Technology Blend & Tools

This is a technology course, so technology is the medium, not an add-on. Two primary tool families are used; the course resists tool sprawl so students go deep, not wide.

• The course IS the technology foundation — every concept is taught hands-on so abstractions become concrete.
• Deliberately no-code: students build real artifacts without programming, so legal computational literacy is universal.
• Tools are kept simple and free where possible, so the focus stays on thinking, not tool-wrangling.

Semester III · Technology Stream · 3 Credits · 0-0-4-0 · Technology-blended (lab-based). Builds on TECH-101/102. No statutory basis. Benchmarked against emerging "AI for lawyers" curricula at JGLS, NLUs and global legal-tech programmes. Industry anchor: LexisNexis/RELX hires "Prompt Engineer — Legal".

1 · Course Overview

Prompt & Context Engineering teaches the skill that turns a generic AI model into a reliable legal assistant: framing the right input. As legal-AI tools proliferate, the differentiating competence is no longer access to the tool but the ability to instruct it precisely, supply the right context, and structure its output for legal work. This course builds that competence systematically.

Students learn prompt patterns, context-window management, retrieval grounding and prompt evaluation, applied to real legal tasks — drafting, summarising, research and review. Critically, the course pairs every technique with verification and a clear sense of where prompting cannot rescue a fundamentally unreliable output. It is hands-on throughout, with a personal, version-controlled prompt library as the running deliverable.

Course Objectives

1. Understand how large language models process prompts and context, and why input design drives output quality.
2. Master structured prompt patterns (role, task, format, examples, constraints) for legal tasks.
3. Engineer context — supplying documents, authorities and grounding — to reduce hallucination.
4. Evaluate and iterate prompts systematically, building a reusable, version-controlled prompt library.
5. Apply prompting responsibly with verification, confidentiality and disclosure discipline.

2 · Course Outcomes, BTL & PO/PSO Mapping

BTL key: L2 Understand · L3 Apply · L4 Analyse · L5 Evaluate.

CO → PO / PSO Correlation (3 High · 2 Medium · 1 Low)

PSO1 = Commercial & transactional acumen. PSO2 = Legal-technology & drafting fluency.

Rationale: PSO2 is High throughout — this is a core legal-AI skill course. PO6 (ethics) is High at CO5 because deciding when an output is reliable is a professional-responsibility judgment. PO9 (communication) peaks at CO6 as the prompt library is documented for others. PO5 (research) is High where prompting drives research tasks.

3 · Detailed Syllabus — Six Modules

Six modules, each mapped consecutively to one Course Outcome (Module 1 → CO1 … Module 6 → CO6). Each module is paired with the lab experiments in Section 9.

4 · Case Studies (Module-Mapped)

Case Study B — The Prompt That Lied Module 3

Student tasks

1. Diagnose the failure: what about the prompt invited fabrication?
2. Redesign the prompt to demand grounded, citable output and to allow 'I don't know'.
3. Re-run and verify every claim in the new output against a primary source.
4. Document the before/after and the general lesson.

Deliverables

• A failure diagnosis.
• A redesigned, grounding-forcing prompt.
• A verification table proving the new output is reliable.

Learning outcomes

• Diagnose prompt failure and hallucination (CO3).
• Engineer grounding to fix it (CO4).
• Internalise verification as non-negotiable.

Case Study C — Grounding the Judgment Summary Module 4

Student tasks

1. Compare a naive (no-context) summary against the real judgment to find the errors.
2. Engineer a grounded prompt that supplies the judgment text properly.
3. Produce a faithful summary with paragraph-level references.
4. Verify the summary against the source and document the improvement.

Deliverables

• A naive-vs-grounded comparison.
• A grounded, referenced judgment summary.
• A verification note.

Learning outcomes

• Engineer context and grounding (CO4).
• Evaluate output reliability (CO5).
• Produce trustworthy legal work-product with AI.

Case Study A — The One-Line Brief Modules 1 & 2

Student tasks

1. List the missing information the one-line brief assumes and decide sensible defaults.
2. Build a structured prompt (role, task, format, constraints, examples) for the NDA draft.
3. Run it and evaluate the draft against what a competent associate would expect.
4. Iterate the prompt once to fix the biggest weakness in the first output.

Deliverables

• A structured prompt with stated assumptions.
• The generated first-draft NDA.
• A before/after note on the single most important prompt improvement.

Learning outcomes

• Apply structured prompt patterns to a real drafting task (CO2).
• Understand why input design drives output quality (CO1).
• Build the habit of making assumptions explicit.

Case Study D — The Library Adoption Pitch Modules 5 & 6

Student tasks

1. Assemble evidence that each library prompt is tested and reliable (benchmark results).
2. Prepare a reliability-and-limits statement for the library as a whole.
3. Pitch the library to the 'practice group', addressing the reliability objection head-on.
4. Capture the group's conditions for adoption and revise the library accordingly.

Deliverables

• A tested, documented prompt library.
• A reliability-and-limits statement.
• An adoption pitch and a revision log responding to the group's conditions.

Learning outcomes

• Construct and defend a version-controlled prompt library (CO6).
• Evaluate output reliability and communicate it (CO5).
• Translate a technical artifact into an accountable, adoptable tool.

5 · Active Learning Methods

Four high-impact, course-relevant ALMs. Trivial methods are excluded in favour of methods that build genuine technical-and-judgment competence.

ALM 1 · Prompt Battle

Two teams attack the same legal task with competing prompts; outputs are scored blind on a rubric — building rapid, evaluative prompt skill. Implementation: task released → teams craft prompts in 20 min → outputs scored blind by peers on accuracy/format/verifiability → winning prompt dissected.

ALM 2 · Hallucination Forensics

Students take a fabricated output and trace exactly which prompt weakness caused it, then prove a fix — the diagnostic core of the course. Implementation: bad output supplied → forensic diagnosis → redesigned prompt → verified proof of fix → debrief.

ALM 3 · Context-Budget Challenge

Given a tight context window and a large document, students must decide what to include to get a faithful answer — building context-engineering judgment. Implementation: large doc + small budget → teams choose chunks → run and compare faithfulness → debrief on what mattered.

ALM 4 · Build-and-Adopt (PBL)

Teams build the prompt library and run an 'adoption review' where another team must use only the library to complete a task — testing real usability. Implementation: build library → hand to another team → they complete a task using only it → adoption report → refine.

6 · Course-Based Team Activities

Specific to this course and distinct from the semester-wide Activity Stream. Teams of 4–5; each builds a lawyering ability with technical-and-verification rigour emphasised.

7 · Resources & MOOCs

Prescribed / Reference Texts

Recommended MOOCs

8 · Global Certifications & National Benchmarking

Relevant Certifications (none forced)

Benchmarking Against Nationally Reputed Universities

Reasoning for missed topics: The ~10–20% not covered is deliberate: model fine-tuning and building AI systems move to TECH-401; data-science methods sit in TECH-202. This course stays focused on the universally-needed skill — instructing and grounding AI for everyday legal tasks — so every student masters it before specialising.

9 · Laboratory Experiments (Technology Course)

As a technology-blended course, this course is delivered through a structured lab. Ten experiments run across the semester in the 2-hour weekly lab slot, each producing an artifact. Conduct: each lab has a pre-lab brief, an in-lab task on a live tool, and a submitted artifact; labs are cumulative and feed the Module-6 capstone; a standard lab rubric applies (below).

Lab Evaluation Rubric (per experiment)

10 · Technology Blend & Tools

This is a technology course, so technology is the medium, not an add-on. Two primary tool families are used; the course resists tool sprawl so students go deep, not wide.

• The course is entirely hands-on with live models; every technique is practised, not just described.
• Every output must be verified against a source — prompting never substitutes for legal verification.
• Confidentiality is built in: students learn what client data may not be entered into third-party tools.

Semester IV · Technology Stream · 3 Credits · 0-0-4-0 · Technology-blended (lab-based). Builds on TECH-101/102/201. The most code-forward course in the stream (gentle Python). No statutory basis. Benchmarked against legal-analytics programmes; feeder for Legal Data Scientist roles (₹10–25 LPA).

1 · Course Overview

Legal Data Science & Predictive Analytics teaches lawyers to work with legal data at scale — to mine large volumes of legal text, find patterns, and build simple predictive and classification tools. As courts digitise and firms accumulate matter data, the ability to ask quantitative questions of legal information is a genuine differentiator and the foundation of the fast-growing legal-analytics field.

This is the most technical course in the stream, but it is taught for lawyers, not engineers: Python is introduced gently and used as a means to legal ends — text mining a corpus of judgments, classifying clauses, analysing judge or court patterns, and visualising findings in a dashboard. Throughout, the course foregrounds the ethical and methodological cautions of legal prediction: correlation is not law, and a model's output is evidence, not a verdict.

Course Objectives

1. Introduce data-science workflow (acquire, clean, analyse, model, communicate) applied to legal data.
2. Build gentle Python fluency for legal text mining and analysis (pandas, spaCy, transformers at a high level).
3. Teach legal-text analytics: classification, extraction and pattern-finding on judgments and contracts.
4. Introduce predictive analytics (outcome prediction, judge/court analytics) with strong methodological caution.
5. Develop the ability to communicate quantitative legal findings honestly through visualisation and dashboards.

2 · Course Outcomes, BTL & PO/PSO Mapping

BTL key: L2 Understand · L3 Apply · L4 Analyse · L5 Evaluate.

CO → PO / PSO Correlation (3 High · 2 Medium · 1 Low)

PSO1 = Commercial & transactional acumen. PSO2 = Legal-technology & drafting fluency.

Rationale: PSO2 High throughout (a data-science course). PO7 (societal/environmental concern) appears because predictive analytics raises fairness and access-to-justice questions. PO6 (ethics) is High at CO5 — the ethics of legal prediction is central. PSO1 rises at CO6 where analytics informs commercial decisions.

3 · Detailed Syllabus — Six Modules

Six modules, each mapped consecutively to one Course Outcome (Module 1 → CO1 … Module 6 → CO6). Each module is paired with the lab experiments in Section 9.

4 · Case Studies (Module-Mapped)

Case Study B — The Clause Classifier Module 3

Student tasks

1. Prepare a labelled sample of clauses (liability vs. not).
2. Build a simple classifier and run it over the corpus.
3. Evaluate precision and recall; find the false negatives.
4. Decide how the tool should be used given its error rate (human review of flags).

Deliverables

• A working clause classifier.
• A precision/recall evaluation.
• A recommendation on safe deployment with human-in-the-loop.

Learning outcomes

• Analyse legal text via classification (CO3).
• Evaluate model validity and error cost (CO5).
• Learn why 'accurate' models still need lawyers.

Case Study C — The Biased Bail Predictor Modules 4 & 5

Student tasks

1. Train/inspect the predictor and measure overall accuracy.
2. Disaggregate performance across groups to reveal disparate impact.
3. Trace the bias to its source (proxy variables, historical data).
4. Write a recommendation: why this model cannot be used for decisions, and what would be required.

Deliverables

• A fairness analysis across groups.
• A root-cause explanation of the bias.
• An ethics memo on (non-)deployment.

Learning outcomes

• Apply analytics with methodological care (CO4).
• Evaluate bias and ethics of legal prediction (CO5; PO6/PO7 High).
• Internalise the limits of data-driven justice.

Case Study A — The Disposal-Time Question Modules 1 & 2

Student tasks