Everyone talks about digital transformation, but it only clicks when you see it in action. One clear example is Intelligent Document Processing: a cloud and AI workflow that goes far beyond OCR to route, extract, and validate business data at scale.
That is the kind of work Suprakash Dutta leads. He is a Senior Solutions Architect at Amazon Web Services, advising large enterprises on how to move from legacy stacks to modern, cloud-native systems. His projects range from rebuilding point of sale platforms across 2,400 stores to migrating complex banking estates at Fujitsu. In this interview, he defines transformation today, explains why architecture choices matter, and shows where generative AI is taking the enterprise.
Suprakash, as an expert in digital transformation, can you tell us what this concept means today? Many interpret it differently. How would you define it now?
Digital transformation has evolved beyond simply digitizing existing processes. Today, it is a fundamental rethink of how an organization delivers value through technology. I define it as the strategic use of digital technologies to significantly improve business performance by transforming operational models, customer experiences, and organizational culture.
What makes digital transformation different now is that it’s not just about implementing technology. It’s about creating new capabilities. Companies aren’t merely replacing paper forms with digital forms; they’re using AI to predict customer needs, using the cloud to launch products globally in minutes, and building data driven decision frameworks that weren’t possible before.
The most successful transformations I’ve seen focus on solving real business challenges rather than chasing technology trends. For example, at a large specialty retailer, we didn’t just modernize their point of sale system. We reimagined how their entire retail operation could function with a cloud native architecture. This created a unified customer experience across their 2,400+ retail locations that their legacy systems couldn’t support. We transformed their business by replacing their outdated Swing based system with a modern, event driven microservices architecture.
I would also like to add that everything we are discussing reflects only my personal views and of course does not represent the position of my employer.
Why is it essential for businesses to transition to modern digital architectures — even those that have been operating successfully for a long time?
Even successful businesses with established processes face the fact that yesterday’s success formula rarely guarantees tomorrow’s relevance. Modern digital architectures offer three critical advantages that legacy systems simply cannot match.
Let’s return to the earlier retailer example. Their legacy Swing based point of sale system was hindering innovation, even though it still worked. The transformation to a cloud native architecture delivered specific benefits they couldn’t achieve before:
First, it enabled unprecedented agility through an API first, event driven microservices architecture using Amazon API Gateway, AWS Lambda, and AWS Fargate. Development teams can now rapidly deploy new features to enhance the shopping experience.
Second, by migrating from Oracle Exadata to Amazon Aurora PostgreSQL, we optimized their data model, reducing 2,600+ tables to just 400 and improving system performance while lowering costs. They also moved to AWS Graviton3 R7g database instances, further improving performance.
Third, the solution enabled business capabilities that weren’t possible before. Their new unified customer data platform built on Amazon DynamoDB and Amazon OpenSearch Service allows them to deliver personalized customer experiences across all channels. This leads to higher customer loyalty and larger transaction values.
You work with some of AWS’s largest clients, essentially acting as an external CTO. Which technologies are primarily transforming business today — cloud, AI, automation?
I believe the primary business transforming technologies driving enterprise innovation today are: cloud, AI (especially generative AI and AI agents), and automation working together. The cloud provides the foundation, AI brings intelligence, automation implements at scale, and generative AI reimagines processes entirely. Working with AWS customers as a Senior Solutions Architect, often acting as an external advisory field CTO, I’ve observed this convergence across industries.
In retail and CPG, for example, generative AI is enabling new customer experiences. Imagine fashion retailers creating virtual shopping assistants that can have natural conversations with customers about style preferences, or virtual try on experiences that generate realistic product visualizations based on customer photos.
AI agents are transforming CPG operations, including inventory management, where they can analyze multiple data sources to better anticipate demand patterns. For instance, a beverage company could use AI agents to simulate various pricing and promotion scenarios across different markets.
When these technologies converge, that’s where the magic happens. Think of a grocery retailer combining cloud based inventory systems with generative AI for personalized customer communications and automated reordering systems, which can reduce waste and improve customer satisfaction.
These examples reflect the kinds of transformations I’m seeing clients explore as they integrate these technologies into their existing ecosystems.
Tell us more about some of the Retail Transformations or Modernization stories. What did this kind of transformation mean for the customer’s business?
The point of sale modernization for Upbound Group addressed critical challenges that held back innovation and customer experience. Upbound was struggling with an outdated point of sale system and three major pain points: technology that was hard to scale, inconsistent customer data across channels, and limited operational visibility. Their monolithic, on premises legacy setup couldn’t meet changing demand, and developers couldn’t build new features quickly.
We designed an API first microservices architecture using Amazon API Gateway, AWS Lambda, and AWS Fargate. Teams can now release new features rapidly, and business teams gained the flexibility to combine products to meet goals.
We also migrated a 45 TB Oracle Exadata database with 2,600+ tables to Amazon Aurora PostgreSQL. Using AWS DMS and AWS Schema Conversion Tool, we optimized the model to 400 tables and under 30 TB. Upbound moved to AWS Graviton3 based R7g database instances, further improving performance.
To enhance customer experience, we implemented a global customer database on Amazon DynamoDB with search powered by the Amazon OpenSearch Service. This created unified customer profiles across all touchpoints, enabling personalized marketing and integrated support.
For business continuity, we deployed a multi-region architecture with pilot light, Amazon Aurora Multi AZ clusters with read replicas, and cross-region replication for Amazon S3 and DynamoDB.
The impact has been transformative. The serverless first approach enables rapid development cycles without infrastructure constraints. Store operations now save hundreds of hours per month across 2,400+ locations, staff focus on strategic work, and customers get a consistent omnichannel experience with personalized interactions and better service quality.
What made this project successful was close collaboration with Upbound’s technical teams, forming two pizza teams across platform engineering, application development, and enterprise services.
During your time at Fujitsu, you oversaw the migration of a large number of banking applications to the cloud. Why is moving to the cloud especially critical for the financial sector?
Financial institutions face a unique combination of challenges that make cloud adoption not just beneficial but essential. During my time at Fujitsu, I helped customers migrate about 150 banking applications, which showed me several key reasons why I believe the cloud is critical for this sector.
First, the financial landscape is increasingly competitive, with fintech disruptors setting new standards for the digital experience. Traditional banks with legacy systems simply cannot match the speed of innovation these nimble competitors offer. The cloud provides banks with the agility to rapidly develop and deploy new features and services, reducing time to market from months to days.
Second, financial institutions face major security challenges. While many executives initially feared cloud security, the reality is that cloud providers invest billions in security measures that most banks cannot match independently. For the legacy bank we worked with, moving to well-architected AWS and Azure landing zones fundamentally improved their security posture, implementing defense in depth strategies, comprehensive encryption, and advanced threat detection.
Third, regulatory compliance is an ever-growing burden in finance. Cloud platforms offer sophisticated governance and compliance tools that automate many aspects of regulatory adherence. Our implementation included automated compliance checks and documentation, significantly reducing the manual effort required for audits.
Finally, financial institutions deal with massive data volumes and unpredictable demand spikes. The cloud’s elastic nature allows them to scale computing resources precisely when needed, for example during month end processing or tax season, rather than maintaining expensive infrastructure sized for peak loads.
This transformation wasn’t just technical; it required cultural change. We worked closely with the bank’s leadership to build cloud fluency across their organization, ensuring the migration delivered lasting business value beyond the technical implementation.
We know you also mentor students at the University of Texas at Dallas. In your opinion, what are young professionals today lacking as they enter the industry? And what, in your view, is truly impossible to achieve without live mentoring?
In my mentoring work with UT Dallas Data Science Club students, I’ve observed that today’s young professionals bring strong technical skills but often show gaps in three areas.
First, systems thinking: understanding how their technical solutions fit into the larger business context. Many students can build sophisticated machine learning models but struggle to explain how they create business value. I encourage them to start with business problems rather than technical solutions, asking “why” before “how.”
Second, effective communication across teams. Technical professionals must explain complex concepts to nontechnical stakeholders and translate business requirements into technical specifications. This skill is often not emphasized in school but is essential for career advancement.
Third, resilience and adaptability. Technology changes rapidly, and many students struggle with ambiguity and continuous learning. I help them understand that being comfortable with constant change is a competitive advantage.
As for what’s impossible without live mentoring, I believe contextual judgment is extremely difficult to develop without guidance. Books and courses can teach principles, but applying them to messy, real world scenarios requires nuanced judgment that comes from experience. For instance, knowing when to pursue technical perfection versus meeting business deadlines, or how to navigate organizational politics while maintaining technical integrity.
How do you personally stay up to date with technologies that evolve every single day? Do you have any personal rituals or $habits when it comes to professional development?
Staying current in technology requires deliberate practice rather than passive consumption. My approach combines structured learning with practical application and knowledge sharing.
I maintain a disciplined “learning workflow” that starts with broad awareness and narrows to deep expertise. Each morning, I spend 20 to 30 minutes scanning technology news sources and my curated Twitter list of thought leaders. This keeps me aware of emerging trends and technologies.
For technologies directly relevant to my work, I dedicate time blocks in my calendar — typically 5 to 7 hours weekly — for focused learning. These are not optional. I treat them as seriously as client meetings. During these blocks, I might complete hands-on labs, build proofs of concept, or work through certification materials.
I use certifications strategically, not just for credentials, to force systematic learning. My 9 AWS certifications, along with Azure, Google Cloud, and Oracle certifications, have each expanded my knowledge into areas I might not have explored otherwise.
But knowledge without application isn’t truly internalized, so I maintain several personal projects where I can apply new technologies in concrete ways. For example, when generative AI was emerging, I built a virtual IT troubleshooting assistant that integrated with multiple enterprise systems, which later became the basis for an AWS ML blog post.
Perhaps most importantly, I teach and write about what I learn. My technical publications, workshop development, and mentoring push me to explain concepts clearly, which deepens my own understanding. As Richard Feynman said, “If you want to master something, teach it.”
Finally, I maintain a network of technology professionals across different specialties. Our regular exchanges — sometimes formal knowledge sharing sessions, sometimes informal conversations — expose me to diverse perspectives and technologies I might otherwise miss.
This systematic approach keeps me current and helps me guide enterprise clients with confidence.
