Embedded Files
Projects
Projects
AI Support Agent for Faster Ticket Resolution
AI Support Agent for Faster Ticket Resolution
Built a full-stack AI Support Agent that automates support ticket operations from triage to draft response.Client goal was to reduce manual effort in reading every ticket, classifying urgency, checking customer context, and writing replies from scratch. Their team needed faster turnaround, consistent response quality, and a workflow that still kept human approval before sending.I developed a Next.js + Supabase dashboard with an AI processing pipeline that:
Built a full-stack AI Support Agent that automates support ticket operations from triage to draft response.Client goal was to reduce manual effort in reading every ticket, classifying urgency, checking customer context, and writing replies from scratch. Their team needed faster turnaround, consistent response quality, and a workflow that still kept human approval before sending.I developed a Next.js + Supabase dashboard with an AI processing pipeline that:
Classifies tickets by category, priority, and sentiment
Classifies tickets by category, priority, and sentiment
Enriches customer context (company type, likely plan, account value)
Enriches customer context (company type, likely plan, account value)
Generates context-aware draft responses with reasoning
Generates context-aware draft responses with reasoning
Supports human-in-the-loop review, editing, and approval
Supports human-in-the-loop review, editing, and approval
Tracks ticket lifecycle through statuses (new, processing, ready, approved, sent)
Tracks ticket lifecycle through statuses (new, processing, ready, approved, sent)
Main challenge was keeping AI outputs structured and reliable for production-style workflow decisions. I solved this with schema-like prompting, validation-safe handling, and clear state transitions in the backend API routes.
Main challenge was keeping AI outputs structured and reliable for production-style workflow decisions. I solved this with schema-like prompting, validation-safe handling, and clear state transitions in the backend API routes.
Result: significantly faster draft response creation, improved consistency across support interactions, and a scalable workflow that can handle larger ticket volume without proportional team expansion.
Result: significantly faster draft response creation, improved consistency across support interactions, and a scalable workflow that can handle larger ticket volume without proportional team expansion.
Note: portfolio media uses safe demo data and anonymized examples.
AI Lead Enrichment & Outreach System
AI Lead Enrichment & Outreach System
Built an AI-powered lead enrichment and outreach automation platform for a B2B agency looking to scale their cold outreach without hiring more SDRs.The Challenge: Client was spending 2+ hours per lead manually researching companies, identifying pain points, and writing personalized emails.
Built an AI-powered lead enrichment and outreach automation platform for a B2B agency looking to scale their cold outreach without hiring more SDRs.The Challenge: Client was spending 2+ hours per lead manually researching companies, identifying pain points, and writing personalized emails.
My Solution:
My Solution:
• Web scraper that automatically pulls company data from websites
• Web scraper that automatically pulls company data from websites
• AI engine that analyzes the business and identifies specific pain points
• AI engine that analyzes the business and identifies specific pain points
• Personalized email generator that writes unique outreach per prospect
• Personalized email generator that writes unique outreach per prospect
• Dashboard to manage, approve, edit, and send emails in bulk
• Dashboard to manage, approve, edit, and send emails in bulk
Result: What took 2 hours per lead now takes 30 seconds. Client can process 50+ leads per day with zero manual research.
Result: What took 2 hours per lead now takes 30 seconds. Client can process 50+ leads per day with zero manual research.
Tech: Next.js, TypeScript, Supabase, Groq AI, web scraping
Tech: Next.js, TypeScript, Supabase, Groq AI, web scraping
AI Lead Enrichment repo: https://github.com/websytdesign/ai-lead-enrichment
Page updated
Google Sites
Report abuse