AI Lead Qualification & CRM Automation
I built a production AI system that enriches, scores, and qualifies leads using Apollo, PDL, and Claude AI — writing decisions and talking points back to HubSpot.
How It's Used:
A Sales Rep creates a HubSpot contact. n8n detects it, enriches via Apollo/PDL, runs Claude AI scoring, writes results back in under 60s. Qualified leads trigger Slack alerts.
Key Highlights
– 5-agent pipeline: Research, Qualification, Routing, Engagement, Feedback
– Parallel enrichment + Claude AI analysis
– n8n: HubSpot trigger → Vercel → CRM writeback + Slack
– Configurable ICP settings — no code needed
Outcomes
✓ End-to-end HubSpot → AI → CRM in under 60s
✓ 4-dimension scoring with feedback loop
✓ Live ops dashboard
Stack - Python · Vercel · Supabase · n8n · Claude API · Apollo · PDL · HubSpot · Slack · React
1
0
25
Jira Service Management Optimization
Redesigned and optimized Jira Service Management for a FAANG client's service operations environment. Rebuilt JSM workflows, SLA rules, and approval logic tied to Azure AD groups. Integrated Microsoft Teams and Outlook for seamless notifications and cross-team communication. Built automation rules for alerts, user flags, and comment restrictions on closed tickets.
Results:
✅ Redesigned SLA rules and approval workflows
✅ Azure AD-based approval logic
✅ Teams & Outlook integration for notifications
✅ Automated alerts reducing manual intervention
✅ Delivered improvements that led to repeat business
Tech: Jira Service Management, Azure AD, Microsoft Teams, Outlook, Automation Rules
0
37
Automated Jira-GitHub Board Integration
Configured a fully automated Jira board integrated with GitHub for Modigie's development team. Built 5 automation rules that transition ticket statuses based on pull request activity — PR opened moves the ticket to In Progress, review requested moves to In Review, and PR merged auto-resolves to Done. Branch naming convention (feature/MOD-###) auto-links commits and PRs to Jira tickets.
Eliminated manual ticket updates entirely, reducing context switching for developers and keeping the board accurate in real-time with a full audit trail of code-to-ticket activity.
Results:
✅ Zero manual Jira updates required
✅ 5 automation rules for full PR lifecycle
✅ Bi-directional GitHub ↔ Jira sync
✅ 100% PR tracking with auto-linked branches
Tech: Jira, GitHub, Jira Automation Rules, Webhooks
0
39
Built a full-stack AI automation system that transforms how the company responds to RFPs.
Users upload RFP documents (PDF/Excel) through a web portal. An n8n pipeline processes the upload: AI parses the document to extract key fields (issuer, scope, budget, deadline), classifies the service line, matches content from an Airtable library (team bios, references, portfolio), and generates 80-90% complete proposal drafts with cover letter, company overview, technical approach, and more. Output is a formatted Word document ready for review.
Results:
✅ Reduced proposal creation from 4-6 hours to under 30 minutes
✅ 90% auto-complete rate on proposal sections
✅ 14+ RFPs processed through the pipeline
✅ Full proposal dashboard with status tracking
Tech: n8n, OpenAI, Airtable, Vercel, Webhooks
4
5
225
Built an AI-powered triage bot (WeSki Bot) for a ski travel company using n8n, Slack, HubSpot, Jira, and GPT-4.
Agents trigger the bot in Slack with a booking ID. It verifies trip details in HubSpot in real-time, then walks through a structured triage flow via threaded replies: collecting issue description, urgency, OPS needs, actions taken, and refund details.
GPT-4 processes inputs and auto-creates a Jira ticket with booking details, triage notes, SLAs, and refund amounts. A Slack notification posts with links to the Jira ticket and HubSpot deal.
Results:
✅ 60% faster response times
✅ Zero manual data entry
✅ 100% lead capture
✅ 24/7 automated triage
Tech: n8n, Slack, HubSpot, Jira, GPT-4, Google Sheets