AI Agents That Work While You Sleep.
Claude AI agents, n8n workflows, and SP-API integrations that eliminate manual operations — giving your team back hours every single week.
Featured Projects(4)
Real work. Real results. Real numbers from real clients.
20+ n8n Automation Workflows
Full Operations Automation Suite — Next Degree Products USA
Designed, built, and deployed 20+ production n8n workflows that fully automated the data operations of a 7-figure Amazon brand — from order ingestion and inventory alerts to automated reporting emails and Slack notifications.
Operations team manually pulled reports from 8 different platforms every morning, sent update emails by hand, and had no real-time alerting for inventory or order issues. This consumed 3–4 hours of analyst time daily with high error risk.
Built a suite of n8n workflows hosted on AWS EC2 to automate: SP-API order ingestion → BigQuery, daily P&L email summaries, inventory low-stock Slack alerts, Keepa price drop notifications, weekly performance digest PDFs, and Sellerboard auto-sync. All workflows monitored with error retry logic.
Claude AI Reporting Agent
LLM-Powered Business Intelligence — Amazon Analytics
Built a Claude AI agent with tool-use capabilities that answers natural language questions about Amazon business performance — pulling live data from BigQuery and generating executive summaries, trend analysis, and anomaly explanations on demand.
Leadership needed instant answers from their data but couldn't write SQL or navigate dashboards. Analysts spent hours fielding ad-hoc data questions via Slack that could take 30–60 minutes each to answer manually.
Deployed Claude claude-opus-4 with a custom tool library connecting to BigQuery: `query_sales_data`, `get_inventory_status`, `analyze_ppc_performance`, `explain_anomaly`. Natural language input → SQL generation → formatted business summary output. Integrated into Slack via webhook so leadership could ask questions directly in their workflow.
OpenAI-Powered Product Listing Generator
Automated Content Creation — Amazon FBA Brand
Built an LLM pipeline using OpenAI GPT-4 that automatically generates SEO-optimised Amazon product listings — title, bullets, description — by analysing Keepa data, competitor listings, and keyword research data from BigQuery.
Brand launched 30–50 new ASINs per quarter. Each listing required 2–3 hours of copywriting plus keyword research. This bottleneck delayed product launches and cost $200–300 per listing in copywriter fees.
Engineered a Python pipeline that pulls top-50 competitor ASINs from Keepa, extracts keyword data from SP-API Search Term Reports, and feeds structured context into GPT-4 with a custom system prompt trained on high-converting Amazon listing patterns. Outputs title, 5 bullets, and description in brand voice. Integrated into Airtable for human review before publishing.
Power Automate + SharePoint Workflow Suite
Internal Operations Automation — Enterprise Client
Built a comprehensive Microsoft Power Automate workflow suite integrating SharePoint, Teams, and Outlook to automate internal approval processes, document management, and weekly reporting for an enterprise operations team.
Approval workflows ran entirely via email chains with no tracking. Documents were stored inconsistently across local drives and email attachments. Weekly reports required manual data aggregation from 4 SharePoint lists.
Deployed 12 Power Automate flows handling: document approval routing with auto-escalation, SharePoint form submissions with Teams notifications, weekly KPI digest from SharePoint lists → formatted Outlook email, and automated folder creation on new project setup. All flows tracked in a central Power BI monitoring dashboard.
Ready to build something like this?
Every project starts with a conversation. Let's talk about your goals and how we can deliver results like these for your business.