Hi, I’mBhargav Vaghasiya
I’m a Liferay Architect who builds AI agents that do real work on Liferay DXP. They read live data, raise requests and route approvals, and Liferay’s own permissions decide what they may do. Behind that are 6+ years of building and modernizing enterprise portals.

- Java
- Liferay DXP
- OSGi
- REST APIs
- GraphQL
- Microservices
- PostgreSQL
- MySQL
- Elasticsearch
- Maven
- Jenkins
- Docker
- Git
- React
- TypeScript
- JavaScript
- Tailwind CSS
- Liferay AI Hub
- Claude Code
- GitHub Copilot
- Cursor
- MCP
- OAuth 2 / SSO
- Linux CLI
- Postman
- IntelliJ IDEA
Agents that do the work
Proofs of concept on Liferay AI Hub. Every answer comes from live Liferay data, and every action passes Liferay’s permissions and a person’s approval.
Portal Assistant
An actionable chatbot that looks up, raises and approves leave, expense and IT requests on live Liferay data. Liferay's permissions decide what it may do, not the prompt.
Grievance Management Desk
An AI Hub agent that takes in grievances and routes each one to the right department.
Email Invoice Intake
Invoices arriving by email are picked up by an agent and queued on a review dashboard for a person to check.
Website Page to Liferay Draft
An agent that reads an existing website page and turns it into a draft Liferay page.

Liferay Architect,
now building AI agents
I build on Liferay DXP, and lately that means AI agents. On Liferay AI Hub I build agents that read live portal data, raise requests and route approvals, with Liferay’s own permissions deciding what they may do. I demoed one, a Portal Assistant for leave, expenses and IT, at the Liferay India User Group Meetup in October 2026.
That rests on 6+ years of enterprise Liferay work: moving portals off 6.1 and 6.2 onto modern DXP without losing what business users depend on, replacing spreadsheet processes with governed onboarding workflows, wiring in SAML, LDAP and CIAM sign-in, and building finance, governance and partner portals, some of them serving several regions.
I start with what the platform already gives you, like Objects, Kaleo workflows and client extensions, and write a custom module only when those run out. I’ve published 3 apps on Liferay Marketplace and took 1st place at the Liferay Marketplace Fragment Hackathon 2026.
Work Process
A structured but flexible process that scales with project size - from breaking down business needs to monitoring a controlled production rollout.
The goal at every step: scope clearly, design for maintainability, and avoid overengineering traps.
1. Requirement Breakdown
Convert business needs into technical modules, identify Liferay objects and extension points, and define scope boundaries early.
2. Architecture Design
Decide modules vs. client extensions and API-first vs. portal-first; design data flow and the permission model upfront for scale.
3. Development
Client Extensions, Objects and APIs on the backend; React-based Client Extensions on the frontend; optimized queries and reusable components.
4. CI/CD Integration
Jenkins pipelines for automated build and deploy flows, with environment-specific configuration handling.
5. Testing & Hardening
Functional and regression testing, security validation of roles and permissions, and performance sanity checks for portal load.
6. Deployment & Monitoring
Controlled production rollout, log monitoring and issue triage, with iterative fixes post-release.
Portfolio
Selected enterprise work across Liferay architecture, client extensions, CI/CD automation, and security.
Licensing Onboarding & Governance Platform
Enterprise platform for governed, multi-stage partner onboarding, with a dozen workflows, over a hundred Liferay objects and a React-driven dashboard.
Meeting & Governance Portal Modernization
Migration of a legacy Liferay 6.2 meeting portal to the latest DXP with a React frontend, calendar integration and the ability to schedule and run conferences.
Headless CMS for Mobile Application
Liferay as a headless CMS serving content to a mobile app over GraphQL, with bulk content management, CIAM login and Elasticsearch tuning for performance.
Curious how these were built? Read the engineering blog.
Tech Stack
The tools and platforms I reach for across backend, frontend, the Liferay ecosystem, AI agents and AI-assisted engineering, infrastructure, and security - grouped by where they do their work.
Core Engineering
6 capabilities- Java (Enterprise Development)
- Liferay DXP / Portal Development
- OSGi Framework
- REST APIs
- Microservices
- SQL Optimization
Liferay Platform
9 capabilities- Liferay DXP 7.x / 2024+ Quarterly Releases
- OSGi Modules & Service Builder
- MVC Portlets
- Client Extensions
- Liferay Objects (Low-Code)
- Headless REST & GraphQL APIs
- Batch Engine (Import / Export)
- Liferay Cloud / Liferay SaaS
- Upgrades & Data Migration
Liferay Content & Experience
10 capabilities- Page Builder & Display Page Templates
- Themes, Fragments & Page Templates
- Web Content, Structures & Templates
- Documents & Media
- Collections & Asset Publisher
- Liferay Forms
- Segmentation & Personalization
- A/B Testing & Analytics Cloud
- Liferay Commerce
- Blogs, Wiki & Message Boards
Liferay AI, Security & Ops
8 capabilities- Liferay AI Hub (Agents & Actionable Chatbots)
- Workflow Engine (Kaleo)
- Search (Elasticsearch / OpenSearch, Blueprints)
- Permissions, Roles & Service Access Policies
- SSO (SAML, OpenID Connect, OAuth 2)
- Publications & Staging
- Multi-Site & Virtual Instances
- Localization & Translation Management
Frontend
6 capabilities- React.js
- TypeScript
- JavaScript (ES6+)
- HTML5 / CSS3
- Tailwind CSS
- Liferay Frontend Client Extensions
AI & Automation
5 capabilities- Liferay AI Hub agent workflows
- Actionable chatbots
- AI-assisted development
- Workflow automation
- Prompt-based engineering for dev productivity
AI-Assisted Engineering
7 capabilities- Claude Code
- GitHub Copilot
- Cursor
- Prompt-to-prototype workflows
- AI pair programming & code review
- MCP tool integrations
- Rapid UI prototyping
DevOps & CI/CD
5 capabilities- Jenkins Pipelines
- Maven
- Git / GitHub
- Docker (basic)
- CI/CD automation
Security & Systems
5 capabilities- Role-Based Access Control (RBAC)
- Authentication & Authorization flows
- Web security testing
- Permission debugging
- Bug bounty mindset
Database & Data Layer
5 capabilities- PostgreSQL / MySQL
- Query Optimization
- Schema Design
- Liferay Objects
- View-based data modeling
Tools
8 capabilities- IntelliJ IDEA
- Eclipse
- Visual Studio Code
- Liferay Developer Studio
- Postman
- Linux CLI
- Docker
- Jira / Agile workflows
Blog
Technical write-ups on Liferay, security, and system design - documenting real migration lessons and debugging workflows.
See these ideas in practice across my project case studies.
What I do?
I build AI agents on Liferay, and the enterprise systems they run on: backend engineering, frontend modernization, DevOps automation and security, designed to scale and stay maintainable.
Whether it’s a first AI Hub agent, a portal upgrade, a CI/CD overhaul or an access-control redesign, I bring a system-level perspective to the whole stack.
Say Hello!