In an interview with TimesTech, Sandeep Malhotra, Chief Strategy, Solutions & AI Officer, Digitide, shared insights on why enterprises are moving toward hybrid AI models that combine automation with human judgement. He discussed how organizations across BFSI and insurance are embedding responsible AI, workflow-driven intelligence, and governance-led frameworks to scale AI adoption while maintaining trust, compliance, and operational resilience in increasingly complex business environments.
Read the full interview here:
TimesTech: You emphasize that the future of enterprise AI is “hybrid” rather than fully automated. How should organizations strike the right balance between AI‑led efficiency and human judgement, especially in high‑stakes environments like insurance and BFSI?
Sandeep: In high‑stakes environments, the objective of AI is not autonomy—it is assured decision‑making at scale. BFSI and insurance workflows are exception‑heavy, regulation‑driven, and often involve nuanced judgment. Fully automated systems perform well in narrow, rules‑based contexts, but they struggle when decisions require interpretation, accountability, or ethical oversight.
The right balance is achieved by assigning AI the role of accelerator, not arbiter. AI should handle scale—pattern detection, prioritisation, anomaly identification, and decision support—while humans retain control over exceptions, approvals, and outcomes with financial or regulatory impact.
This hybrid model preserves trust, ensures accountability, and allows AI systems to continuously improve through human feedback. It is not a temporary compromise—it is the operating equilibrium enterprises are adopting to scale AI responsibly.
TimesTech: You describe AI as a lifecycle capability rather than just a tool. Could you walk us through how enterprises should approach AI across strategy, engineering, operations, and governance to unlock sustained value?
Sandeep: Enterprises that succeed with AI treat it as infrastructure, not experimentation. That requires a lifecycle approach across four tightly connected layers.
First is strategy, where organisations define clear business outcomes, value pools, and risk boundaries. Without this, AI initiatives fragment into disconnected pilots. Also, organization clearly understand the human and technology readiness to embrace AI, since future of work is hybrid, it also initiate the right change management strategies and lay foundation for effective governance
Second is workflow‑centric engineering. AI must be embedded into core processes—not layered on top. This means defining new learning models, creating agentic workflows, defining human‑AI interaction models, and ensuring interoperability with existing systems.
Third is operations, where AI systems are continuously monitored, recalibrated, and optimised as data patterns, regulations, and market conditions change.
Finally, governance and responsible AI must be built into the architecture from day one—covering explainability, security, auditability, and compliance. Governance cannot be retrofitted at scale.
When approached this way, AI transitions from innovation to enterprise capability.
TimesTech: From your experience, what are the key lenses—industry, process, and persona—that truly determine AI ROI, and how do priorities differ across stakeholders like CIOs, CFOs, and COOs?
Sandeep: AI ROI is contextual and must be evaluated through three lenses.
Industry defines the value logic. In BFSI and insurance, value concentrates around risk accuracy, regulatory compliance, customer lifetime value, and operational resilience—not just cost reduction.
Process determines scalability. AI compounds value fastest in high‑volume, decision‑intensive workflows. Automating tasks yields incremental gains; setting up agentic workflows delivers structural impact.
Persona ensures executive alignment.
- CIOs prioritise system stability, integration, and scalability.
- CFOs focus on economic returns, risk mitigation, and predictability.
- COOs care about throughput, cycle time, and operational consistency.
As an AI-First organization, our personas has clear goal on future of work – hybrid, human in loop and AI-first
AI initiatives that align with board‑level KPIs scale faster and deliver measurable ROI.
TimesTech: Digitide is working closely with insurers through platforms like IIP and AI‑powered data hubs. Where are you seeing the fastest compounding value of AI across insurance workflows, and what differentiates successful implementations from failed “quick wins”?
Sandeep: In insurance, AI value compounds fastest in decision‑dense workflows—underwriting, claims adjudication, risk assessment, parametric modelling, fraud detection, and policy servicing. These areas generate continuous data loops, allowing AI models to learn, adapt, and improve outcomes over time.
What differentiates success from failure is workflow architecture.
Quick wins automate isolated tasks—document classification or chat interfaces. They validate potential but rarely change economics.
Successful implementations redesign workflows end‑to‑end—how cases are triaged, how exceptions are routed, how decisions are explained, and how accountability is embedded. When AI reshapes decision chains rather than tasks, value compounds instead of plateauing.
TimesTech: Responsible AI is becoming a critical conversation globally. How are enterprises operationalizing governance, explainability, and compliance, and why do you believe regulated sectors are setting the benchmark for AI adoption?
Sandeep: Regulated sectors are setting the benchmark because they cannot scale AI without trust. In BFSI and healthcare, explainability, auditability, and compliance are prerequisites—not enhancements.
Leading enterprises are embedding governance into system design—defining autonomy thresholds, human‑in‑the‑loop controls, bias monitoring, and explainability directly within workflows. Compliance informs model behaviour, not just post‑deployment reporting.
Responsible AI does not slow innovation. It accelerates adoption by reducing uncertainty for regulators, customers, and internal stakeholders. Trust is what allows AI to move from pilot to platform.
TimesTech: With AI innovation increasingly driven by ecosystems rather than standalone efforts, how is Digitide enabling partnership‑led, services‑driven innovation, and what does this shift mean for enterprises building scalable AI capabilities today?
Sandeep: Enterprise AI is too complex to build in isolation. Digitide operates a services‑led, ecosystem‑driven model, where hyperscalers, technology partners, and domain platforms come together around business outcomes. Pulse.ai, our overarching AI-framework, assets and capabilities on the backbone of AI-incubation centre, collaborative innovation at scale is foundational to our success.
Our role is orchestration—bringing industry context, process depth, and execution rigor to translate technology into impact. Innovation is co‑created with customers through structured MVP frameworks, AI labs, and flexible engagement models.
The shift from product‑led to services‑driven AI means enterprises can scale faster, reduce risk, and embed AI directly into operations—rather than managing fragmented tools. That is what enables AI to become durable, governed, and customer‑centric at scale.
