I'm Nilesh Mohite — Principal Consultant at Xebia and an AI-native software architect with 18+ years across .NET, Azure, and cloud-native systems. I design platforms where microservices and AI agents work as one architecture — proven in e-commerce, healthcare, banking, and insurance.
I started in .NET in 2008, when service-oriented architecture was the frontier, and I've carried enterprise systems through every shift since — into the cloud, into containers, into event-driven microservices, and now into the agentic era.
Along the way I've worked on teams that shipped treasury and insurance systems for leading Indian financial institutions, FHIR-compliant healthcare platforms, a global film-pricing tool for a major entertainment studio, and a global medical-device manufacturer's production microservices. Today, as a Principal Consultant at Xebia, I lead development on a large event-driven commerce platform on Azure — and build AI agents where Claude acts as a reliable component inside well-governed systems. Not demos: production software.
Outside client work I build in public — open-source agents, multi-agent experiments with BMAD, and writing about what actually works when AI meets enterprise constraints.
Leading a microservices, event-based e-commerce platform for a large enterprise program on Azure — and building autonomous AI agents with the Claude API and BMAD multi-agent framework.
Designed .NET 6 microservices for a health-benefits catalog platform using CQRS with per-service PostgreSQL schemas, on Azure with API Management, Service Bus, Redis, and SignalR.
Led support for a global medical-device manufacturer's distributed microservices systems — .NET Core, NServiceBus, Elasticsearch, SAP Hybris, OpenID.
Designed a FHIR-standards healthcare API system and a real-time QA analytics platform on Azure — with security via Azure AD, OAuth 2.0, and RSA encryption.
Built microservices for a Fortune 500 technology distributor and a global film-pricing tool for a major entertainment studio on Azure PaaS, with Okta authentication for Fortune 50 clients.
Delivered a Centralized Accounting System for a leading private-sector bank's treasury and SOA services for a major life-insurance provider.
ERP, financial accounting, and inventory systems on .NET, WPF, and SQL Server — where the foundations were laid.
How I structure large systems so they stay maintainable.
Applied with intent, not by habit.
The core toolset for services and front ends.
The top 20 Azure services I reach for — compute, data, integration, security, and delivery.
Agentic systems built for production, not demos.
The delivery pipeline behind every release.
Relational, document, cache, and search — chosen per workload.
Industry platforms integrated in production, across regulated and commercial domains.
A personal open-source project: Ticket2PR bridges Jira and GitHub. When a ticket moves to "In Progress," a webhook triggers Claude to generate the code, commit it to a new branch, and open a pull request — no manual intervention. Three interchangeable generation modes — a single-shot Claude service, an iterative agentic tool-use loop, and a mock service for local testing — with HMAC-SHA256 webhook validation and credentials kept out of source control.
// live replay: an autonomous .NET agent reads a Jira story, writes the code, and opens the PR. I built it. Source on GitHub ->
Part of a large enterprise program delivering a microservices-driven, event-based e-commerce platform, where I lead development across several .NET 8 domain services on Azure AKS, integrated with a mix of best-of-breed commerce, marketing, and customer-experience SaaS platforms — provisioned with Terraform and Azure DevOps CI/CD.
Contributed to the design of a highly scalable, standards-compliant healthcare REST API system on FHIR, built with .NET Core and Azure — Functions, Key Vault, SQL Database, and Storage — secured with Azure AD, OAuth 2.0, and RSA asymmetric encryption. Within the same team I led the real-time QA analytics workstream, with an Angular / D3.js front end on AKS and SignalR.
Worked within the team that delivered a VOD/EST pricing tool for a global film & entertainment studio, supporting major digital channel partners worldwide — ASP.NET Core and Angular on Azure PaaS with Okta authentication. Also contributed microservices for a Fortune 500 technology distributor.
Part of the teams behind a Centralized Accounting System for the treasury division of a leading private-sector bank, a risk-assessment and cost-benefit reporting platform, and SOA services for a major life-insurance provider across Dev, UAT, and Production — the regulated-industry foundation behind how I design auditable systems today.
Compute/storage separation, identity federation, messaging, orchestration — the primitives outlive any vendor. I architect on primitives so clouds become deployment details.
Agents are only as trustworthy as their constraints. I treat guardrails, tool-permission design, and multi-agent orchestration as first-class architecture, not afterthoughts.
FHIR standards, OAuth 2.0, RSA encryption, auditable treasury systems — in healthcare, banking and insurance, compliance is a design constraint from day one. I've shipped under it.
You will come across Nilesh Mohite as a very mature, level headed and knowledgeable person. I worked with him in my last project for around 1.5 year. He is one such technical colleague who keeps himself updated and upgraded with every change in technology relating to his field as well outside it. A calm and composed employee who puts forward his views and suggestions quantitatively leaving no space for questioning his answer/knowledge. He is an asset for every team. I personally am impressed with his time management skills. The way he manages his work and up-scales his technical and professional skills is commendable. Extremely knowledgeable in Azure, .Net and Cloud computing. He is a great asset for any team.
Nilesh is a good person in nature. I have been working with him since last 2 years. He has got ample of knowledge in the field of Dot net development. He keep himself updated to new technologies and his skills are not only restricted to Dot.net but also knows the other technologies…
All stories live at niluniverse.com ↗
How one table and one transaction absorb broker outages, retries, and traffic spikes so your microservices don't have to — the pattern that never trends, yet quietly holds up production.
A request died somewhere across forty services — no body, no witness. Three witnesses (logs, metrics, traces), one thread (the trace ID), and the waterfall exhibit that cracks the case in one screen.
You've hired a brilliant improv actor for a role that must be identical eight shows a week. Schemas, validators, a deterministic orchestration sandwich, and evals — with a live interactive demo.
