AWS & Azure Cloud and Data Architect
9+ years · Agentic AI Engineering
I’m a forward deployed engineer and AI implementation lead, and I design and build cloud platforms that other teams and agents can rely on.
I use agentic AI in day-to-day delivery: multi-agent spec-writing playbooks for data-pipeline, web and infrastructure work, with adversarial verification and an ADR loop; harness and context engineering; and agent-runnable tooling for Azure DevOps backlog sync and workspace setup. My knowledge base on AI engineering covers six disciplines: Prompt Craft, Context, Harness, Intent, Specification and Orchestration Engineering.
Claude Code
I created the security-first, zero-trust API framework every new web application on AWS starts from, taking a new service’s setup from 3 weeks to 2 days, and led the CI/CD and Infrastructure as Code modules (AWS CDK, Terraform on Azure) now used by nearly 20 services.
I architected the analytical data product on the medallion lakehouse: Kinesis ingestion, AWS Glue (PySpark) ETL, Apache Iceberg on Amazon S3 Tables and Lake Formation tenant isolation, all in AWS CDK. It cut report time for 100,000 records from 3 minutes to under a minute.
I'm an AWS & Azure Cloud and Data Architect based in Leeds, United Kingdom, and today an Agentic AI Implementation engineer, Forward deploying AI solutions to business.
I design and build cloud platforms that other teams and agents can rely on. Over eight years I've moved from .NET web applications and microservice migrations on Azure to leading AWS cloud and data architecture. My recent work includes CI/CD and reusable infrastructure modules, platform standards for APIs and caching, GDPR compliance, and an analytical data product on the medallion architecture. I also use agentic AI in day-to-day delivery: multi-agent spec-writing playbooks for data-pipeline, web and infrastructure work, with adversarial verification and an ADR loop; harness and context engineering; agent-runnable tooling for Azure DevOps backlog sync and workspace setup; and a structured knowledge base on AI engineering disciplines.
Answers composed ahead of time from his CV facts and project write-ups. Each one cites where it came from.
Anything else? Pass holders can ask their own assistant, through the site’s connector.
Kumar's portfolio says his current role is Senior Software Engineer at PebblePad since September 2023. Its flagship projects are Agentic Development, Analytical Data Product, API Framework and Build & Deploy.
Kumar's portfolio says he has over 8 years of experience.
Kumar's portfolio lists four roles: Senior Software Engineer at PebblePad since September 2023; Senior Product and Platform Engineer at Accenture, July 2021 to August 2022; Senior Software Engineer at Imaginea Technologies (now part of Accenture), December 2017 to June 2021; Programmer Analyst Trainee at Cognizant, February 2017 to November 2017.
Kumar's portfolio lists these projects with agents: Agent Memory & Knowledge (Developer): Hand-built agent loops, harnesses and workflows, with memory and documentation served to agents through RAG and an MCP server; Agentic Development (Lead): Agentic work follows the platform's own infrastructure and deployment rules, and data services move from specification to tested code through one repeatable workflow, delivered 3–5x faster.
Kumar's portfolio lists these data projects on AWS: Analytical Data Product (Architect): One centralised analytical foundation for reporting and self-service analysis, separated from the operational systems and built entirely on AWS. It cut the report time for 100,000 records from 3 minutes to under a minute; Storage Limitation (Developer): The platform captures and keeps less personal data, and holds it only as long as it is needed, in compliance with the GDPR.
Kumar's portfolio lists ten certifications held: AWS Certified Machine Learning Engineer – Associate, October 2026; AWS Certified Data Engineer – Associate, June 2026; AWS Certified Solutions Architect – Associate, February 2026; Microsoft Certified: Azure Solutions Architect Expert, July 2021; Microsoft Certified: DevOps Engineer Expert, February 2022; Microsoft Certified: Azure Developer Associate, May 2021; Microsoft Certified Solution Associate, Web Applications, September 2017; Microsoft Certified Professional, Programming in C#; Microsoft Certified Professional, Querying Data with Transact-SQL, January 2018; Microsoft Technology Associate, Software Development Fundamentals, December 2016. It lists two more in preparation: Claude Certified Architect – Foundations; Claude Certified Architect – Professional.
Kumar's portfolio says he mentored 11 engineers and 5 apprentices at Imaginea Technologies (now part of Accenture) and trained two senior and three junior developers at PebblePad.
Kumar's portfolio does not take messages. His contact details are on the About panel.
Good architecture is mostly invisible. It's what lets everything else move fast and keep working under load.
From cloud and data architecture to the platforms and web apps on top, I work across the whole stack. I design systems, then build, ship and run them.
I design well-architected systems on AWS, Azure or anywhere, turning requirements into resilient, scalable and cost-aware solutions.
Multi-agent spec-writing playbooks with adversarial verification and an ADR loop, built on harness and context engineering.
Data modelling, pipelines and analytics, through to operationalising machine learning from features to production.
Infrastructure as code, CI/CD and Kubernetes, and the .NET APIs and web apps on top. I ship and run platforms, not just build them.
MSc Advanced Computer Science (Cloud Computing)
2022 – 2023
B.Tech in Computer Science Engineering
2012 – 2016
9+ years of professional experience across 4 roles
Leading agentic development and AWS cloud and data architecture: multi-agent workflows that deliver data services 3–5x faster, the analytical data product, the API framework, and the CI/CD and infrastructure modules behind nearly 20 services.
Architected and built microservice platforms on Azure, with their data pipelines and stores, delivered through infrastructure as code (Terraform) and CI/CD.
Modernised legacy monoliths into cloud-hosted microservices on .NET Core, designing APIs, data models and integrations and building analytics and reporting features, and mentored 11 developers and 5 apprentices.
Built ASP.NET web forms and MVC applications, and contributed to proof-of-concept work across data access, services and early analytics.
A decade of delivery and research, from architecting Azure platforms to modernising decade-old monoliths and prototyping with AI. Each card opens into the context, my contribution and the outcome.
Built agent tools and workflows that carry the platform's infrastructure and deployment rules from specification to tested code.
Architected an analytical data product on the medallion architecture, built entirely from AWS services.
Built the security-first, zero-trust framework for web applications on AWS, from Cognito with PKCE and API Gateway authorizers to WAF, KMS and PrivateLink, and built APIs on it.
Led the CI/CD for AWS cloud-native development and the plug-and-play infrastructure modules it deploys.
Built agent tools and workflows that carry the platform's infrastructure and deployment rules from specification to tested code.
Architected an analytical data product on the medallion architecture, built entirely from AWS services.
Built the security-first, zero-trust framework for web applications on AWS, from Cognito with PKCE and API Gateway authorizers to WAF, KMS and PrivateLink, and built APIs on it.
Led the CI/CD for AWS cloud-native development and the plug-and-play infrastructure modules it deploys.
Hand-built agent loops, harnesses and workflows, with memory and documentation served to agents through RAG and an MCP server.
Designed and built this site with spec-driven, agentic development, serverless on AWS.
Made the platform safe to forget: tombstoning that keeps audit trails, proper masking, and controlled removal of data beyond its retention limit.
Introduced Valkey as a distributed, multi-layer and content-aware cache that lets the platform scale horizontally.
Built an event-driven compliance job on AWS Lambda and EventBridge so the company's compliance checks always run.
Built a centralised document-generation platform on Azure, validating policy data with InRule across products.
Remapped a legacy system into modern APIs and mentored the team through the transition.
Migrated a monolith to microservices on AKS with dynamic, reflection-driven reporting.
Architected an Angular + GraphQL skills-search platform for managers, deployed on AKS.
Stood up a buffer microservice that kept forms and documents flowing correctly across systems.
Worked on modernising AS400 and in-house legacy systems into a microservice architecture for quoting, payment, policy and renewals.
Led the modernisation of a decade-old, 42-project WebForms application into microservices.
Enabled seamless, secure single sign-on transfer between a .NET and a Java platform.
Refactored a reimbursement monolith into microservices with stateful, automated email workflows.
Built the Angular UI skeleton and helped shape the GraphQL backend over a normalised schema.
Prototyped across a broad stack and built custom UiPath activities for ImageRight automation.
Built the multi-tenant PostgreSQL gold layer of the analytical data product, where one EF Core migration set provisions every tenant across all three isolation strategies.
Built a complete in-house analytical data platform on PostgreSQL: a poor man's data warehouse, evolved through nine versions.
Authored a .NET library that evolves domain models over time while keeping full backward compatibility with older stored data.
Authored a .NET library that intercepts deletions and stores immutable archive snapshots with TTL policies and cascade archiving.
Built a reusable Terraform sample library covering the Azure estate, consumed by hundreds of pipeline-driven deployments.
Prototyped an access-review tool that surfaces the data reviewers need to decide.
Built a smart PDF-form prototype that prefills from prior data and generates completed forms.
Prototyped a sign-language interpreter using Custom Vision and text-to-speech.
Whenever you want to achieve something, keep your eyes open, concentrate and make sure you know exactly what it is you want. To accomplish great things we must not only act, but also dream, not only plan, but also believe.
Machine learning in production
October 2026Operationalising machine learning on AWS: preparing data and engineering features; building, training, tuning, deploying and monitoring models in production; and applying MLOps, automation, model governance, security and cost control, centred on Amazon SageMaker.
Data pipelines & analytics
June 2026Building and operating data pipelines on AWS: ingesting, transforming and orchestrating batch and streaming data; designing and managing data stores; cataloguing, governance and security; and optimising data for analytics with services such as Glue, Kinesis, Redshift, Athena and Lake Formation.
Well-Architected solution design
February 2026Designing and deploying well-architected solutions on AWS: resilient, high-performing, secure and cost-optimised architectures across compute, storage, networking and databases, applying the AWS Well-Architected Framework to turn business requirements into systems that scale.
Azure Solutions Architect Expert
July 2021Expertise crafting cloud and hybrid solutions on Azure across Compute, Network, Storage, Monitoring and Security, advising stakeholders and translating business requirements into solutions aligned with the Azure Well-Architected Framework, then implementing them alongside Developers, Administrators, Security and Data Engineers.
DevOps Engineer Expert
February 2022Bridging development and operations to enable continuous delivery: collaboration, code management, infrastructure provisioning, source control, security, compliance, CI, testing, delivery, monitoring and feedback, with a strong focus on Terraform, Azure DevOps and GitHub.
Azure Developer Associate
May 2021Responsible across all stages of development (requirements, design, deployment, security, maintenance, performance tuning and monitoring), spanning SDKs, storage, data connections, APIs, app auth, container deployment and debugging.
Web Applications
September 2017Professional developer using Visual Studio and ASP.NET (incl. .NET Core) to architect web solutions: designing and developing ASP.NET MVC applications, planning user-interaction solutions and deploying to multi-tier environments including Azure.
Programming in C#
CertifiedWriting applications in C#: managing program flow, events and asynchronous processing; creating and using types, encapsulation and inheritance; debugging, validating input, encryption and diagnostics; and implementing data access with collections, LINQ, files, serialization and databases.
Querying Data with Transact-SQL
January 2018Managing data with Transact-SQL: SELECT queries, joins, functions and aggregation, data modification, subqueries, APPLY, table expressions, pivoting, temporal and non-relational data, plus stored procedures, functions, triggers, views, error handling and transactions.
Software Development Fundamentals
December 2016Core programming principles: storage and data types, decision structures, repetition and error handling; object-oriented concepts; application life cycle; algorithms and data structures; web, desktop and database application fundamentals.
Building production applications with Claude
In preparationArchitecting production applications on Claude: where the model fits in a system and what it can and cannot do, prompt and context design, reliable integration through the Claude API, evaluation, and the guardrails that make a deployment safe to run.
Enterprise-scale Claude deployments
In preparationDesigning and leading enterprise-scale Claude deployments: solution and integration architecture, context engineering, evaluation and optimisation at scale, and AI governance, risk management and responsible deployment practices across stakeholders.
Notes on cloud & data architecture, .NET engineering and the practice of shipping software.
A chat window bolted onto a product is not an AI strategy, and neither is an assistant licence mandated for developers - the case for adoption with intent, and the staged-maturity evidence behind it.
The diagnosis came back genuine and the repository is empty - now the discipline the old system used to impose for free must be rebuilt by hand. Greenfield delivery as three questions (decide, prove, ship) and the gates an AI-assisted team still owes itself.
Teams ask for a rewrite as though greenfield were a reward for enduring the old codebase - it is a diagnosis with a test attached, and misdiagnosis is how rewrites fail. Part 1 of 2 on choosing the field you build on.
Pick the one closest to the role. Each is two pages and carries the same facts.
Applied AI
Agentic AI & Cloud Engineer · multi-agent systems, RAG and MCP on AWS
Cloud & .NET
C#, .NET 10 and ASP.NET Core microservices on AWS and Azure · Terraform & AWS CDK
Data Engineering
AWS lakehouse · Kinesis, Glue (PySpark), Apache Iceberg, Lake Formation
Everything
All three domains, with the standout project from each