AI Agent Builder & Automation Engineer
I started in web development, then discovered AI and automation in early 2025 — and went all in. In under a year, I've shipped a live AI voice agent handling 2,000+ real business calls, engineered multi-stage data pipelines for US eCommerce brands, built Python automations that run silently in the background every single day, and delivered AI audits for companies across insurance, furniture, and retail. Every project solves a real operational problem — no fluff, no demos that don't ship.
Completed multiple internships and took on freelance clients in the web development space — building and delivering real projects, meeting client briefs, and learning to work independently without hand-holding. This is where I developed the discipline of shipping things that actually work.
My first major AI project — and it went straight to production. Built a fully autonomous AI voice agent for a US business that handles their entire inbound call flow: answers questions from the company's own knowledge base, books appointments, collects project details, routes information to the right team member, and logs complaints. No human needed for routine calls. It has handled over 2,000 real calls and continues to run today.
Designed and deployed a fully hands-off social media engine for a realtor client. Every day, the system automatically generates a fresh AI image and matching caption, then publishes it to their social accounts — zero manual effort required. What used to take 20–30 minutes of daily work now runs completely on autopilot.
Database Overhaul: Walked into a database that was barely functional — riddled with duplicate entries, misnamed fields, and broken variables accumulated over years of patchy data entry. Audited the entire structure, built targeted automation scripts to catch and resolve inconsistencies at scale, and finished what automation couldn't with careful manual review. Took it from unusable to reliable.
Supplier Quote Intake Pipeline: The company was receiving dozens of supplier quotes from China daily — arriving as PDFs, Excel files, and CSVs in wildly inconsistent formats. Built an intelligent extraction pipeline that reads each incoming email, identifies the correct product codes regardless of formatting variation, parses the quote data accurately, and stores it cleanly in the database. Significant engineering effort went into handling the unpredictability of real-world supplier documents.
Automated Order Sync (SellerLegends → Database): Built a production-grade Python automation hosted on GitHub Actions that wakes up every morning, logs into SellerLegends, downloads the latest Amazon order data, and syncs it into the company's internal database — keeping everything current without anyone lifting a finger. Extended the same architecture to handle a second daily report serving a separate operational purpose.
Live Tariff Monitoring System: US–China tariffs change frequently and unpredictably — and for an Amazon seller sourcing from China, a missed tariff update directly hits margins. Built an automated monitoring system that watches for tariff changes across the company's product catalog, verifies updates, and writes them to the database in real time. One less thing the team needs to manually track.
Ad Creative & New Product Launches: Produced a range of AI-generated ad creatives for new product lines — experimenting with different visual styles, angles, and audience personas to build out launch-ready creative libraries without the cost of a full creative team.
TikTok UGC Automation (In Progress): Architected the early-stage infrastructure for a multi-account TikTok content engine — designed to generate and post UGC-style product videos at volume across multiple accounts to drive organic reach and top-of-funnel awareness.
Brought in by Highlane Partners to build a bespoke AI-powered image workflow platform for a furniture brand. The brief was complex: users needed to upload product photography, direct AI edits to the images, review and approve outputs, and feed approved assets into downstream workflows — all in one system. Delivered a cohesive, working solution that handled the full creative review cycle from upload to approval.
Partnered with an independent consultant to conduct a thorough AI readiness audit for an insurance company — mapping their current operations, identifying where AI could create meaningful efficiency gains, and designing a practical implementation roadmap. Delivered an architecture centered on 3 core AI agents, each targeting a distinct operational bottleneck, with supporting automation workflows connecting them. Gave the client a clear, actionable picture of where and how to adopt AI — not just a report, but a buildable plan.