Featured Projects: Automation and AI Systems I Built and Run

-

This is a living index of the systems I'm most proud of — each one is in daily production use, and each links to a deeper technical write-up. If you're a recruiter or hiring manager, this page is the fastest way to see what I actually build.

1. AI Email-Triage Pipeline (LLM classification in production)

A pipeline that classifies hundreds of ticketing emails per day (presales, releases, cancellations) using multimodal LLMs, then distributes summarized reports automatically to the relevant business departments. I fine-tuned GPT-4o-mini on a labelled dataset I built from PDF archives, put OpenAI and Gemini behind a single interchangeable JSON contract for cost/performance trade-offs, and wrote a defensive C# parsing layer — balanced-brace JSON extraction, tolerant custom JsonConverters, code-enforced business invariants — so a single malformed model response never breaks the batch.

Stack: C#/.NET, OpenAI API (fine-tuning), Gemini, IMAP, SQL Server
Read more: Classifying Event-Ticket Emails with AI

2. .NET Automation Hub (10+ system integrations, ~25 scheduled jobs)

A unified automation hub for a ticket-resale business that integrates 10+ external systems — POS, banking APIs, Gmail, Slack, Google Sheets — into roughly 25 scheduled jobs, eliminating hours of daily manual operations work. It includes event-driven automation with Hangfire background jobs plus workflow automations built with Power Automate, Azure Logic Apps, n8n, and Python, deployed on AWS EC2 and Azure.

Stack: .NET, Entity Framework, Hangfire, Serilog, Google Sheets API, AWS EC2, Azure
Read more: Building an Automation Hub for a Ticket-Resale Business in .NET

3. Domain Research SaaS Platform (multi-tenant, SQL at scale)

A domain research platform built over four months, featuring a feature-entitlement system for a multi-tenant model that gates pages, API endpoints, and individual data fields by subscription tier across the full stack. The hardest engineering problem was SQL performance: I cut a synchronization procedure's I/O roughly 10x by diagnosing a 13GB clustered-index scan over 130M+ row tables and designing targeted covering and filtered indexes.

Stack: React, .NET Core, RESTful APIs, SQL Server
Read more: Four Months Building a Domain Research Platform

4. Serverless Portfolio Analytics API (AI-generated commentary)

A serverless analytics API for a wealth-management firm: Azure Functions on Python 3.11, deployed through GitHub Actions CI/CD, with OpenAI integration generating analysis commentary over Monte Carlo simulations and time-series forecasting (Prophet, ARIMA).

Stack: Azure Functions, Python, OpenAI API, Prophet, ARIMA, GitHub Actions

5. Core-Card Repayment Gateway at HSBC (100K → 1M daily transactions)

Before my current automation work, I implemented a core-card repayment gateway (Java Spring Boot, Vue.js) in eight months at HSBC and scaled it from 100K to 1M daily transactions through database tuning and horizontal service scaling — work that earned a CEO Recognition Award. Operating 24/7 statutory payment channels in a regulated, audit-driven banking environment shaped how I think about reliability in everything I build now.

Stack: Java Spring Boot, Vue.js, Oracle, Jenkins CI

Want the full picture? See the About page or reach me at letanphp@gmail.com. I'm open to backend, platform, and AI-infrastructure roles.

Comments

Post a Comment

Popular posts from this blog

AWS API gateway, S3

-
Image
  Tutorial: Create a REST API as an Amazon S3 proxy in API Gateway - Amazon API Gateway Using REST API to upload files to the S3 bucket - BlueGrid : BlueGrid AIM role Edit Trust Relationship {   "Version": "2012-10-17",   "Statement": [     {       "Effect": "Allow",       "Principal": {         "Service": "s3.amazonaws.com"       },       "Action": "sts:AssumeRole"     },     {       "Effect": "Allow",       "Principal": {         "Service": "apigateway.amazonaws.com"       },       "Action": "sts:AssumeRole"     }   ] } Next, go to API Gateway --> create a REST API. Create the folder & item resource create GET -->  with "tanldttestingapigateway" is a S3 bucket name. /{folder}/{item} - GET - Method Execution --> to add URL path parameters Create the fo...
Read more

Business case: Monitor mailbox and auto-save the attachment to a SharePoint

-
Image
Business Case: - Monitors a shared mailbox for messages that contain the words "expense report" - Saves the attachment to a SharePoint Library called ExpenseReports - Marks the email as Read - Marks the email as Complete  Solution: Enter a value for Flow name, and then search the triggers for When a new email arrives in a shared mailbox (V2): Select SharePoint site Mark as read Flag email Result:
Read more