Indegene Limited

rs,

INDGN/SE/2024-25/129 March 26, 2025 BSE Limited, Phiroze Jeejeebhoy Towers, Dalal Street, Mumbai- 400001, India. Scrip Code: 544172 Dear Sir / Madam, Sub: Investor Pres

3x

& Experience with Trust & Global Scale Powered by 3P (Precision, Personalization, Performance) 3x Variation in Creative Personalization Asset creation in real time 45% Hyper Automation in Replic

45%

onalization, Performance) 3x Variation in Creative Personalization Asset creation in real time 45% Hyper Automation in Replication More Impactful Content Hyper Personalization (n=1) Real-time en

60%

ase In Point: Contextualizing core capabilities for Aggregate Report sections to deliver up to 40-60% automation PBRER Sections (as archetype report) Effort reduction (after QC of Gen. AI output)

65%

o RSI Clinical and non-clinical exposure | Use in special population and other post auth use 50-65% Input format standardization drives higher automation 30-45% Input format standardization drive

50%

ch as medication error/ lack of efficacy | Overview of signals and signal and risk evaluation 40-50% Literature Characterization of risk and benefits 40-55% 30-45% • • • • • • • • • • Relev

55%

gnals and signal and risk evaluation 40-50% Literature Characterization of risk and benefits 40-55% 30-45% • • • • • • • • • • Relevant info. from contributions (Gen. AI + programmatic autom

96%

formats • Summarization of information within in-scope PSUR report sections Outcomes delivered 96% Quality and consistency (pre-medical writer’s review) >50% Effort reduction 2x Speed Solu

2x

delivered 96% Quality and consistency (pre-medical writer’s review) >50% Effort reduction 2x Speed Solution Highlights Indegene formed a cross-functional team for development, deployment, a

40%

s for appropriateness, integrity and plagiarism Improves content quality when used upstream | 30-40% reviewer productivity | Reduces oversight errors and brings review standardization © 2025 Indegen

70%

eserved. 42 NEXT Literature Surveillance NEXT Literature Surveillance is envisioned to automate ~70% of literature processing effort Literature Sources Automated/ scheduled searches De-duplication

Initial Phase

1

Intermediate Phase

1

Current Phase

1

Enhancement

1

Regulatory Alignment

1

Current Phase

Agents and Multi-Agent Orchestration • Zero-shot, few-shot: Basic interaction • Responses grounded in curated external • Autonomous, reasoning-capable agents with foundational models using explicit prompts. • • Limitations: Inconsistent outputs, prone to hallucinations knowledge sources. Limitations: Limited active reasoning capability collaborating in real-time Enhanced accuracy and contextual reliability Tools, memory, and reasoning support • • Increasing model capabilities, modalities, access, guardrails, and performance + maturing interoperability standards Current Challenges Domain Gap Lack of deep domain- specific context Validation Issues Difficulty validating agent output and decision logic Change Management Complexity integrating agentic-workflows ROI Realization Challenges in use-cases and reducing time to true ROI Regulatory Concerns Compliance and risk management complexity © 2025 Indegene. All rights reserved. 2 Our PoV on Building Agentic Business Applications Current Tre

Regulatory Alignment

Direct mapping to compliance and governance – auditable traceability Knowledge First, Agents Follow: Expert-curated domain knowledge graphs augmenting business process agents leads to shorter time to ROI. © 2025 Indegene. All rights reserved. 3 True Verticalization Approach © 2025 Indegene. All rights reserved. 4 SME Workbench: Engineering using DSL (Domain Specific Language) Revised SDLC with Domain Experts owning GenAI services and Engineering consuming those services © 2025 Indegene. All rights reserved. 5 Cortex Designed to scale the adoption of LLM-based verticalized agents across use cases with a pragmatic enterprise-grade approach Knowledge Engineering Led By Domain Experts Business Applications (Use Cases) Built With Agents Safety DB Domain Knowledge Graphs LLMs Domain Knowledg e Experiential Feedback Fit-for-Purpose Agents Multi-Agent Orchestration (MAO) API Model Context Protocol (MCP) Enterprise System Integrations (AWS, Azure, SF, OpenAI, etc…) © 2025 Indegene. All rights r