I Build Automations That Stop Manual Data Entry,
Reduce Errors,
And Fill Your Pipeline
Working Systems
Three commercial-grade automations driving revenue today. Eight more in the archive below.
Featured — Commercial Winners
B2B Outbound Prospecting Engine
AP Invoice Reconciliation Engine
Process-First
Engineering Approach
The best automation starts with understanding the business process, not choosing a tool. Six stages from discovery to handover.
Discovery
Understand the existing business process, tools, team structure, and operational bottlenecks through stakeholder interviews and system audits.
Process Analysis
Map the current workflow, identify repetitive work, data quality issues, integration gaps, and failure points that impact revenue or efficiency.
System Design
Define workflow logic, data requirements, integration architecture, error handling strategy, and human approval gates. Produce technical specification.
Development
Build end-to-end solution in n8n/Make/Clay with proper validation, normalization, fallback logic, and comprehensive test scenarios.
Reliability Hardening
Add input validation, exception handling, monitoring, alerting, human approval gates, idempotency keys, and dead-letter queues.
Handover & Documentation
Deliver complete workflow documentation, operational runbooks, troubleshooting guides, and knowledge transfer sessions for the team.
Engagement Model
Remote-First
Asynchronous communication with scheduled overlap for your timezone
Project-Based
Fixed-scope pilots, contract work, or longer-term collaboration
International
Experienced with cross-border teams, Wise/Payoneer/Deel payments
UTC+8 (Philippines)
Overlap with APAC, EMEA, and US West Coast business hours
Engineering Philosophy
"An automation workflow should be designed for the conditions in which it can fail, not only the conditions in which it succeeds."
Process First
Understand the business bottleneck before choosing a tool. Tool selection follows process analysis.
Data Integrity
Validate and normalize at the ingestion boundary. Never trust external payloads.
Design for Failure
External APIs, services, and workflows will fail. Build observability and recovery in from day one.
Human Judgment
Important decisions (outbound, financial) require approval gates. Automation augments, doesn't replace.
Observable by Default
Errors are visible, traceable, and actionable. Centralized logging with debug links.
Evidence Over Claims
Implementation artifacts, test scenarios, and recorded demos matter more than impressive terminology.
Production-Grade
Engineering Stack
Self-hosted on Oracle Cloud Infrastructure (OCI) with Docker, Caddy, and Tailscale. Built for validation, fault tolerance, and observable recovery—no black-box templates.
Infrastructure & DevOps
8 technologiesAI Agents & MCP
6 technologiesAutomation & Orchestration
4 technologiesScraping & Enrichment
5 technologiesBackend, Database & APIs
3 technologiesProgramming & Frontend
4 technologiesCRM & Business Systems
2 technologiesArchitecture at a Glance
Measurable
Business Impact
Modeled impact scenarios and tested pipeline benchmarks from end-to-end system builds.
Lead Sanitization & Scoring
Modeled Benchmark: Replaces repetitive manual CRM cleanup with automated ingestion, bot filtering, and idempotent HubSpot routing.
Built an end-to-end inbound lead pipeline that captures every submission, normalizes phone numbers to E.164, filters bots with behavioral analysis, scores lead quality with AI, and pushes only clean, qualified records into HubSpot automatically with idempotency keys.
E-Commerce Revenue Recovery
Modeled Benchmark: Tracks abandoned checkouts with timed delay gates and Supabase purchase checks before sending personalized follow-ups.
Implemented a Make-based automation with stateful delay gates that tracks cart abandonment, applies smart delays, verifies purchase status before each follow-up, generates AI-written recovery emails, and sends via Gmail only when the customer hasn't purchased. Zero duplicate nudges to converted customers.
B2B Outbound at Scale
Modeled Benchmark: Automates company discovery, scraping, AI personalization, and Slack human-approval gates before rate-limited dispatch.
Deployed an autonomous outbound system that discovers target companies via Apify, scrapes intelligence, writes hyper-personalized outreach emails with AI, routes for human approval via Slack, and dispatches with rate limiting. Research time dropped from 15 minutes to under 30 seconds per prospect. Full compliance logging included.
Technical
Skill Stack
20+ production-grade skills across 5 categories. Hover any skill for details, filter by category, or search.
AUTAutomation3 skills
FROFrontend6 skills
AI/AI/ML4 skills
DATDatabase3 skills
DEVDevOps3 skills
Contact Methods
Let's Fix Your
Daily Operations
Interested in improving lead operations, connecting business applications, or reducing repetitive manual processes? I'd be glad to discuss the operational problem and explore whether automation is the right solution.