Indian banks and NBFCs are adopting a 'forward deployed engineer' model to fix stalled AI projects. By embedding specialists directly into workflows, firms aim to turn AI pilots into working tools. For investors, this shift signals a move toward higher accountability in technology spending, which could improve operational efficiency for financial institutions and change revenue models for IT service providers.
What Happened
Financial institutions in India are changing how they approach Artificial Intelligence (AI) implementation. Many banks, Non-Banking Financial Companies (NBFCs), and Global Capability Centers (GCCs) have struggled to move AI projects past the pilot stage. To fix this, these firms are now adopting the 'forward deployed engineer' model, a strategy that involves embedding senior engineers directly into the business units they serve.
Unlike traditional IT roles where teams work in isolation, these engineers function at the intersection of product management, engineering, and banking operations. Their primary goal is to ensure that AI tools, such as automated credit underwriting or fraud detection systems, actually work within the constraints of real-world banking workflows, regulations, and legacy systems.
Why AI Pilots Often Stall
For investors, the key issue is that technology spending in finance often fails to generate returns if the software cannot be integrated into daily operations. A major bottleneck is the lack of 'context.' An AI model might be accurate in a test environment but may fail in a branch or contact center because it lacks access to the right customer data or conflicts with existing compliance policies.
When these integration gaps occur, AI projects often remain unused. This leads to wasted capital expenditure on software that never achieves scale. The forward deployed model is a direct response to this failure, placing engineers in the field to identify and fix these integration roadblocks in real-time.
Impact on Business and Efficiency
This transition marks a shift in how financial firms measure the success of their tech teams. Previously, success might have been measured simply by completing a project or delivering code. Under the new model, success is tied to 'adoption' and 'operational outcomes.'
If banks and NBFCs can successfully implement this model, it may lead to better operational efficiency. Effective AI implementation can reduce the time taken for credit decisions or lower fraud losses. For investors, this means that banks may eventually see better cost-to-income ratios, though the initial cost of hiring or training these specialized engineers is significant.
Opportunities and Risks for IT Providers
For IT service firms and consulting companies, this creates a new service delivery opportunity. Instead of traditional project-based billing, where a company pays for a set amount of work, this model encourages long-term partnerships focused on capability transfer.
However, there are risks. This approach requires talent that understands both complex coding and financial regulations. There is a risk that companies may struggle to find or afford such specialized talent, leading to cost overruns. Furthermore, any integration effort involves the risk of data security lapses or regulatory pushback if the AI systems do not strictly follow compliance norms.
What Investors Should Track Next
Investors may watch how financial institutions adjust their IT budgets to accommodate this more expensive, but potentially more effective, talent strategy. Key monitorables include:
- Technology ROI: Whether banks report improved metrics in customer service or credit risk after adopting these embedded models.
- IT Service Contracts: Whether large IT providers shift their contracts to focus more on 'outcomes' rather than just 'hours worked' or 'project delivery.'
- Talent Costs: Whether wage inflation in the tech sector, specifically for high-end AI roles, pressures the margins of financial service firms.
- Regulatory Updates: How regulators respond to increased AI usage in core banking functions, as compliance requirements are the biggest barrier to integrating AI into live financial systems.