Ownership:

Full technical ownership of 4 production platforms — Purchase & ERP, Finance Management, HRMS (Facial Recognition + Attendance + Payroll), and Vendor Management — all running on Java 17, Spring Boot, GCP.

Key Engineering Contributions:

• 🏗️ Architected microservices with high-throughput REST APIs and API Gateway routing
• 🤖 Integrated Doc-AI / LLM APIs achieving 92% reduction in manual data entry
• ⚡ Designed Express Purchase workflow cutting inventory onboarding by 80%
• 🌏 Built Malaysian E-Invoicing compliance module eliminating 80% manual invoice effort
• 🎭 Engineered facial recognition attendance with geo-fencing and automated payroll
• 🔍 Deployed Elasticsearch for ledger search achieving ~50ms avg response
• 📨 Implemented fault-tolerant async workflows using GCP Pub/Sub with retries & idempotency
• 🐳 Set up CI/CD pipelines (GitHub Actions, Jenkins), Docker, Linux servers on GCP
• 👨‍🏫 Mentored junior engineers via code reviews and pair programming
• 📊 Optimised PostgreSQL schemas and database queries for production workloads

Java 17 Spring Boot Spring Security Spring Cloud GCP PostgreSQL Elasticsearch Redis Pub/Sub Docker CI/CD AI/LLM APIs Microservices API Gateway BigQuery

• Completed 2,160+ hours of intensive full-stack and backend engineering training
• Developed 3+ production-style applications including Naukri clone, e-commerce platforms, and an Automated Auction System with real-time bidding workflows
• Strengthened core DSA and system design skills applied directly in subsequent engineering roles

Java Spring Boot Hibernate MySQL JavaScript HTML/CSS System Design DSA

• Delivered 3+ feature modules for the Fiery product (C++, SQL), contributing to two stable production releases
• Partnered with QA and business stakeholders for requirements gathering, defect resolution, and release validation
• Built a strong foundation in enterprise software delivery that informed subsequent backend engineering work

C++ SQL Enterprise Software QA Agile

The Problem: Enterprise procurement drowning in manual data entry — invoices, POs, receipts all processed by hand. Inventory onboarding took days. No inter-company automation existed.

What I Built: A full-lifecycle AI-powered procurement intelligence platform covering the entire purchase cycle from request to payment — with LLM at its core.

Core Modules Engineered:

• 📋 Purchase Requisition → Purchase Order — automated approval workflows with configurable hierarchies
• 📦 GRN (Goods Receipt Note) — quantity matching, quality checks, discrepancy handling
• 💳 Billing & Payments — auto-reconciliation with vendor invoices, partial payment support
• 🔄 Returns & Replacements — reverse logistics with automated credit notes
• 🏢 Inter-Company Workflows — auto-creation of PO → SO → GRN across entities
• ⚡ Express Purchase — single-screen fast-track workflow for urgent inventory
• 🤖 Doc-AI + LLM Integration — automated extraction from invoices, POs, receipts
• 🌏 Malaysian E-Invoicing — government-compliant electronic invoicing

Impact Metrics:

📉 Manual data entry    ████████████░  92% reduced
🚀 Onboarding speed     ████████████░  80% faster
🎯 Doc-AI accuracy      █████████████  95% precision
🌏 Invoice effort       ████████████░  80% eliminated
📊 PO processing        █████████████  Real-time
🔄 Inter-co workflows   █████████████  Fully automated

Architecture:

┌──────────┐    ┌──────────┐    ┌──────────┐
│ API GW   │───▶│ Services │───▶│PostgreSQL│
└──────────┘    └────┬─────┘    └──────────┘
                     │
              ┌──────┴──────┐
              │  GCP Pub/Sub │
              └──────┬──────┘
                     │
              ┌──────┴──────┐
              │ Doc-AI + LLM│
              │  Pipelines  │
              └─────────────┘

Stack:

Java 17 Spring Boot GCP Doc-AI LLM APIs PostgreSQL Pub/Sub Docker

The Problem: Finance teams relied on costly third-party tools (Zoho Books, Tally). Reconciliation was manual and error-prone. Searching through thousands of ledger entries was painfully slow.

What I Built: A Zoho Books / Tally-inspired scalable finance platform from scratch — designed for real-time accounting, semantic search, and automated bank reconciliation.

Core Modules Engineered:

• 🏦 Bank Account Management — multi-bank support with real-time balance tracking
• 📒 Payable Ledgers — vendor-wise ledger with aging analysis and payment scheduling
• 💳 Purchase Accounting — end-to-end accounting workflow linked to ERP modules
• 🔄 Real-Time Reconciliation — automated matching of bank transactions with invoices
• 🔍 Elasticsearch-Powered Search — instant ledger search at ~50ms avg response
• 📊 Financial Reporting — real-time dashboards for cash flow, payables, receivables
• 🤖 Doc-AI Bank Statements — automated extraction from bank statements, 95% accuracy
• 💳 Loan & Credit Card Workflows — tracking EMIs, interest calculations, card reconciliation

Impact Metrics:

🔍 Search response      █████████████  ~50ms avg
📊 Reconciliation       █████████████  Real-time
🔄 Auto-matching        ████████████░  95% match rate
🏦 Bank formats         █████████████  Multi-format AI
📒 Ledger entries       █████████████  Millions handled
🎯 Statement accuracy   █████████████  95% Doc-AI

Architecture:

┌──────────┐    ┌──────────────┐
│ REST API │───▶│ Spring Cloud │
└──────────┘    └──────┬───────┘
                       │
          ┌────────────┼────────────┐
          ▼            ▼            ▼
   ┌───────────┐ ┌──────────┐ ┌────────┐
   │Elastic    │ │PostgreSQL│ │ Redis  │
   │Search     │ │          │ │        │
   └───────────┘ └──────────┘ └────────┘
          │
          ▼
   ┌───────────┐
   │ BigQuery  │
   │ Analytics │
   └───────────┘

Stack:

Spring Cloud Elasticsearch Redis PostgreSQL BigQuery GCP Doc-AI

The Problem: Manual attendance tracking was unreliable — buddy punching, proxy attendance, no way to verify field employees. Payroll calculation required hours of HR manual work every month.

What I Built: An intelligent HR attendance + payroll automation platform using facial biometrics and GPS geo-fencing — completely eliminating manual HR overhead.

Core Modules Engineered:

• 🎭 Facial Recognition Check-In/Out — real-time identity verification via computer vision
• 📍 GPS Geo-Fencing — location-based attendance policies (~10m precision)
• ⏰ Shift Management — configurable shift schedules with rotation support and overtime tracking
• 💰 Automated LOP Calculation — leave balances, attendance gaps auto-computed
• 📊 Real-Time Attendance Dashboards — live status of all employees across locations
• 💸 AI Payroll Engine — automated salary computation with LOP deductions, overtime, compliance
• 📱 Multi-Device Support — mobile (field employees) and desktop (office check-in)

Impact Metrics:

🎭 Face recognition     █████████████  Real-time verify
📍 Geo-fence precision  █████████████  ~10m accuracy
💰 Payroll calc         █████████████  100% automated
📊 Dashboards           █████████████  Live real-time
🚫 Proxy attendance     █████████████  Eliminated
⏰ HR manual effort     █████████████  Near-zero

Architecture:

┌───────────┐   ┌───────────────┐
│ Mobile/Web│──▶│ Spring Boot   │
└───────────┘   │ + Security    │
                └───────┬───────┘
                        │
           ┌────────────┼──────────┐
           ▼            ▼          ▼
    ┌──────────┐  ┌────────┐  ┌────────┐
    │ OpenCV / │  │ MySQL  │  │Pub/Sub │
    │ Face AI  │  │        │  │        │
    └──────────┘  └────────┘  └────────┘

Stack:

Java Spring Security OpenCV MySQL GCP Pub/Sub REST APIs Geo-fencing

The Problem: Vendor onboarding was a paperwork nightmare — manual GST verification, PAN checks, bank validation. Compliance tracking was reactive. Terms of trade managed in spreadsheets.

What I Built: A centralized vendor onboarding & compliance automation platform with real-time government API integrations and configurable business rules engine.

Core Modules Engineered:

• ✅ Automated GST Verification — real-time GST number validation via government APIs
• 🆔 PAN Validation — instant PAN card verification through third-party integrations
• 🏦 Bank Account Verification — automated bank account validation before vendor activation
• 📋 Document Compliance Tracking — auto-alerts for expiring licenses, certificates, insurance
• ⚙️ Configurable ToT Engine — item-level pricing rules, return policies, delivery constraints
• 🔗 API Gateway Integration — centralized routing for 5+ government and third-party APIs
• 🚚 3PL Vendor Support — third-party logistics vendor management with SLA tracking
• 🔐 OAuth 2.0 + RBAC — role-based access with multi-level approval workflows

Impact Metrics:

✅ GST/PAN verification █████████████  100% automated
📄 Doc compliance       █████████████  Real-time alerts
🔗 API integrations     █████████████  5+ gov APIs
⚙️ ToT engine           █████████████  Fully configurable
🚫 Manual verification  █████████████  Eliminated
🔐 Security             █████████████  OAuth 2.0 + RBAC

Architecture:

┌──────────┐    ┌─────────────┐   ┌─────────────┐
│ Frontend │───▶│ API Gateway │──▶│Microservices│
└──────────┘    └──────┬──────┘   └──────┬──────┘
                       │                 │
              ┌────────┴────────┐  ┌─────┴─────┐
              │ Gov APIs (GST,  │  │  MongoDB   │
              │ PAN, Bank)      │  │  + Docker  │
              └─────────────────┘  └───────────┘

Stack:

Spring Boot API Gateway OAuth 2.0 MongoDB Docker Microservices Spring Cloud