AI Transformation in Practice: Insights from India’s Consulting Leaders

Abstract

The panel examined how the two largest consulting firms in India are integrating generative AI into their own operations and client‑facing services. Speakers described concrete internal use‑cases (audit‑confirmation automation, AI‑driven simulators, rapid‑campaign generation for MSMEs), the strategic re‑thinking of the classic “pyramid” consulting model, workforce reskilling, and the limits posed by change‑management, data‑governance and pricing pressure. The discussion then turned to broader industry concerns – adoption challenges in large enterprises, competitive threats from pure‑tech firms, the role of AI in GovTech, and the future skill set required for students and professionals. Audience questions broadened the conversation to government‑level AI deployments, the impact on education, and the potential for new Indian AI‑driven unicorns.

Detailed Summary

Moderator (Vedika Kant) opened the session, noting the limited time and inviting the panelists to share concrete internal AI use‑cases.
The central question: “What does AI mean for you internally?” – focusing on workflows, tangible examples, and early‑stage impact.

2. Deloitte’s Perspective – Re‑imagining Consulting (5:00‑20:00)

2.1 Business‑model Inversion

Romal Shetty highlighted the traditional “pyramid” consulting model (1 senior client → 10‑person team).
Generative AI enables an inverted model: 10 clients per AI‑augmented consultant, where ≈80 % of repetitive work is performed by machines and ≈20 % by humans.
This inversion opens the MSME segment (≈75 million firms) – previously inaccessible due to cost – by scaling human‑machine hybrids.

2.2 Workflow Acceleration & Automation

Example: Audit confirmation process – up to 60 k balance confirmations per quarter. Deloitte built a practitioner‑crafted tool that automates these requests, saving ~60 k hours and freeing staff for judgment‑heavy tasks.
Tech‑driven consulting – rapid generation of technical opinions via Gen‑AI; a tax‑focused “Navigate Tax Hub” emerged from internal experimentation, launched 6‑7 months earlier than projected.

2.3 AI‑enhanced Simulation & Design

Deloitte created digital twins for a major automobile plant in Karnataka; AI‑driven simulation identified robot‑clash and material‑flow issues that would have prevented the target cycle time (2 min 32 s).
The same approach is being transferred to hospital ICU layout and Jaguar jet flight simulators—projects that traditionally required aerospace‑level expertise, completed within 40 days thanks to AI‑accelerated modeling.

2.4 Human‑in‑the‑Loop & Governance

Emphasised that AI must remain human‑led, warning against unguarded automation which could generate “serious challenges”.

3. PwC’s Perspective – Platform Adoption & Culture (20:00‑38:00)

3.1 Massive Investment & Upskilling

Sanjeev Krishan noted PwC’s ≈US$1 billion AI commitment in 2023, partnered with a leading hyperscaler, and substantial upskilling budget for staff.

3.2 Democratized AI Tools

Introduced “Chat PwC” – an internal conversational AI platform accessible to all personnel for efficiency gains (document drafting, data extraction, preliminary analysis).
Real‑world impact: a tax‑automation tool (Navigate Tax Hub) stemmed from frontline employee ideas; the product was piloted for 12‑15 months before market launch, underscoring bottom‑up innovation.

3.3 Workforce Shape & Pyramid Evolution

Acknowledged that middle‑management may contract, while junior talent will increasingly focus on critical‑thinking, judgement, and empathy alongside AI‑assisted execution.
Discussed the “cylinder” model – a flexible staffing shape that varies by sector and competency, rather than a one‑size‑pyramid.

3.4 Change Management & Adoption Challenges

Stressed that people resist change; the real hurdle is integration and change‑management, not the technology itself.

4. Cross‑Panel Themes – Common Challenges & Opportunities (38:00‑55:00)

Theme	Deloitte View	PwC View
Adoption & Change Management	Pilots succeed; scaling fails due to human resistance.	Same – internal uptake via “Chat PwC” but governance remains key.
Data Governance & IP	Example of aerospace vendor leaking designs via ChatGPT – data‑security risk.	Not explicitly mentioned, but implied need for secure AI environments.
Token Economics & Cost Shock	Anticipates a future “build‑shock” as AI services move from subsidised to priced models.	No direct comment, but acknowledges cost considerations in upskilling and platform investment.
Regulatory & Industry‑Specific Concerns	SMEs can leap‑frog regulatory burdens; large enterprises need careful compliance.	Emphasises “human‑in‑the‑loop” to meet regulatory standards.
Partnerships with AI Vendors	Already partnered with Harvey (OpenAI‑funded) for tax and legal work; exploring Anthropic collaborations.	Not detailed, but investment in hyperscaler partnership suggests similar ecosystem development.

5. Audience Q&A – Extending the Conversation (55:00‑1:45:00)

5.1 Disruption & Pricing Pressure

Romal: AI commoditisation threatens low‑skill consulting. Example – a furniture‑design startup that failed despite AI‑driven sentiment analysis.
Sanjeev: Pricing must evolve; value‑based billing replaces time‑and‑material. AI can move up the value chain (strategic insight, simulation, transformation) while low‑margin tasks become automated.

5.2 Threat from Pure‑Tech Firms (OpenAI, Anthropic, etc.)

Sanjeev: Consulting firms must partner, not compete, with AI vendors. PwC already works with Harvey; future collaborations with Anthropic are underway.
Romal: Technology is an enabler, not a replacement; consultancy still adds domain expertise, trade knowledge, and regulatory insight that pure‑tech firms lack.

5.3 GovTech & Public‑Sector AI

Romal: Government projects (e.g., road‑cost estimation via geospatial AI, MSME credit‑risk analytics) represent massive opportunities.
Sanjeev: Emphasised scale – public‑sector data can fuel AI‑driven solutions that improve efficiency and transparency.

5.4 Talent Development & Education

Romal: Future skills = critical thinking, judgment, empathy, and AI‑orchestration.
Sanjeev: Need for curriculum overhaul from engineering basics to AI‑augmented problem solving.
Audience (students, educators): Calls for practical AI‑enabled learning and a shift from rote memorisation to concept‑driven, AI‑assisted education.

5.5 AI‑driven Entrepreneurship & Indian Unicorns

Romal: While early AI unicorns are US‑centric, India will eventually produce large AI companies, but capital and ecosystem still heavily US‑biased.
Sanjeev: AI is an irreversible trend; Indian firms must carve own pathways, leveraging the scale of the domestic market.

5.6 Market Outlook & Re‑rating of AI Investments

Audience (Sudhakar Gandhi) asked about possible re‑rating of AI‑heavy companies.
Romal: Some firms will under‑perform, others will thrive; investment cycles will adjust but the overall AI market will continue to expand.

5.7 SME Focus & Open‑Source LLMs

Romal: SMEs can leapfrog using open‑source or niche LLMs; they are agile and not bound by heavy regulatory compliance.
Sanjeev: Emphasised choice of technology stack – no single LLM solves all problems; a portfolio approach is required.

5.8 MarTech for Rapid Campaigns

Audience (Pius, Digivency): Request for a tool that creates instant market‑fit campaigns for SMEs.
Romal: Sentiment analysis and AI‑driven market‑demand detection can power such tools; AI democratizes MarTech for small businesses.

6. Closing Remarks (1:45:00‑1:50:00)

Moderator thanked the panel for “honest” reflections on AI‑driven change.
Panel reiterated that consulting models are resilient but must continuously adapt, leveraging AI as a productivity amplifier rather than a replacement.

Key Takeaways

AI as Business‑Model Inverter – Both firms are shifting from a 1‑to‑10 consulting pyramid to a 10‑to‑1 AI‑augmented model, unlocking the massive MSME market.
Automation of High‑Volume Tasks – Deloitte’s audit‑confirmation tool and PwC’s “Chat PwC” illustrate tangible hour‑savings (tens of thousands of hours) that free staff for higher‑value judgment work.
Human‑in‑the‑Loop Remains Critical – Despite automation, ethical, regulatory, and quality safeguards require expert oversight.
Change Management, Not Technology, Is the Biggest Barrier – Staff resistance, data‑governance concerns, and token‑cost dynamics impede scaling of pilots.
Workforce Reskilling Toward Judgment & Empathy – Middle management may shrink; junior talent will need strong critical‑thinking, empathy, and AI‑orchestration capabilities.
Pricing Pressures Will Push Toward Value‑Based Billing – As low‑skill tasks become commodified, firms must move up the value chain (strategic insight, simulation, transformation).
Strategic Partnerships Over Competition – Both firms are collaborating with AI vendors (Harvey, Anthropic) rather than trying to build proprietary LLMs.
GovTech & Public‑Sector AI Offer Massive Scale – AI can optimize infrastructure costs, improve credit‑risk assessment, and streamline government service delivery.
Education Needs a Paradigm Shift – Future curricula must focus on critical thinking, AI‑augmented problem solving, and interdisciplinary skills (psychology, sociology).
SMEs Can Leapfrog Large Enterprises – With open‑source LLMs and rapid‑deployment tools, SMEs can adopt AI faster, leveling the competitive playing field.

These insights collectively illustrate how India’s leading consulting houses are leveraging AI to reinvent internal processes, expand market reach, and redefine talent development, while grappling with the universal challenges of change management, data governance, and evolving business models.

See Also:

India AI Impact Summit 2026

Explorer

ai-transformation-in-practice-insights-from-indias-consulting-leaders