Muhammad Irsyad - AI Automation | Contra
Work by Muhammad Irsyad
Sign Up
Post a job
Sign Up
Log In
Muhammad Irsyad
AI & Workflow Automation Architect | Founder@OrchFlow Studio
Message
Follow
Profile in progress
Muhammad is building their profile!
Followed by
Maxime V
,
Milan G
, and
Hannes A
Pontianak, Indonesia
Work
Posts
Services
About
Pontianak, Indonesia
0
Automated Inbound Lead Triage & Multi-CRM Sync Engine A low-latency automation pipeline that ingests, scores, and routes qualified B2B leads to CRM and Slack in under 10 seconds. Overview & Problem High-ticket B2B agencies and consultancies often lose qualified leads due to response latency and operational friction. When applications arrive from paid campaigns or booking forms, manual data entry and delayed sales notifications cause warm prospects to go cold. Solution Architecture I engineered a resilient, automated lead triage engine powered by self-hosted n8n: Webhook Ingestion: Captures form submissions instantaneously. AI-Powered Lead Scoring: Validates corporate email addresses, filters out low-intent submissions, assigns an objective 1–10 priority score, and produces a 2-sentence executive summary. Multi-Destination Sync: Clean data is logged immediately to the primary CRM (Google Sheets / Airtable / HubSpot). Real-Time Sales Alerts: Pushes high-scoring leads directly into the team's Slack channel with full context so closers can engage within minutes. Fault-Tolerant Routing: Isolates spam or unqualified submissions to prevent primary database clutter. Results & Impact • Sub-10-second end-to-end processing time. • Zero manual copy-pasting for account managers. • Immediate visibility for high-value prospects.
0
27
1
Lead Qualification Bot: AI-Powered BANT Scoring System (n8n + Gemini) Problem: Manual lead triage wastes sales time and delays response to high-intent leads. I built and stress-tested this system to prove how AI can qualify and route leads instantly — this is a self-directed portfolio build, not a live client engagement, but every scenario below reflects a production-ready workflow. Solution: A form trigger captures lead data, Google Gemini scores the lead using BANT criteria (Budget, Authority, Need, Timeline), and a Code node parses the structured AI output. From there, three branches run in parallel: every lead logs to Google Sheets, every lead gets an automatic Gmail reply, and hot leads specifically trigger an instant Slack alert to the sales team. Result: Tested across 5 distinct lead scenarios (hot/warm/cold) with 100% correct BANT classification. Hot leads trigger Slack alerts within seconds of submission — no manual triage required. The parallel branch architecture means logging, email response, and team alerting all happen simultaneously, not sequentially.
1
168
Challenges
Lovable Challenge
Rive Halloween Challenge
The invideo Editor Challenge
Challenges
Lovable Challenge
Rive Halloween Challenge
The invideo Editor Challenge