Client type: Accounting and financial services firm
Stack: Node.js, Express API, PostgreSQL, Vue.js, OAuth, PDF parsing

Challenge:
The firm used spreadsheets and email to manage client filings. No central system, no status tracking, no automation of recurring financial actions.

Solution:
• Built secure web app for client + internal users
• Developed custom API backend for document intake and workflow automation
• Added role-based login, file uploads, audit logging and PDF parsing
• Integrated with external tax API for automated submissions

Outcome:
• Replaced 18 spreadsheets and email threads
• Reduced filing time from 2 days to 25 minutes
• Full audit trail for compliance
• 99.8% automated document handling

Role: Lead software engineer & system architect
Status: Actively used by 5 internal teams

Client type: International fitness & coaching company
Stack: Flutter (iOS + Android), Firebase, Stripe, REST API, Cloud Functions

Challenge:
Client needed a subscription-based mobile app for workouts, coach scheduling, and in-app video content — delivered fast without maintaining separate iOS and Android codebases.

Solution:
• Developed full Flutter app with shared codebase
• Integrated Stripe for subscription payments
• Built Firebase auth + Firestore data layer
• Added push notifications, video streaming, and progress tracking
• Connected to existing backend through REST API

Outcome:
• 1 codebase → 2 production apps (App Store + Play Store)
• 40% lower development cost vs native apps

Role: Full-stack mobile developer (Flutter + backend)
Status: Published and maintained

Client type: Fast-growing SaaS company
Stack: Make.com, Pipedrive, Stripe, Slack, Webhooks

Challenge:
Sales, finance and support teams worked in silos. Deals were updated manually, invoices were created by hand, and payment status wasn’t synced back into the CRM.

Solution:
Automated the full sales pipeline:
• Pipedrive stage change auto-creates Stripe invoice
• Payment confirmation syncs back into CRM
• Slack sends real-time alerts to finance and support
• Error-handling layer for failed payments and retries

Outcome:
• 0 manual invoicing
• 100% CRM ↔ finance sync accuracy
• 15+ hours saved every week
• Fully automated for 2,000+ transactions

Role: Solution architect & implementer
Status: Live and scaling

Client type: Sustainability consulting firm
Stack: n8n, OpenAI GPT-4, Airtable, Slack API

Challenge:
The client received large batches of inbound leads but had no automated way to qualify, enrich or route them. Manual screening took 2 hours per batch and slowed response time.

Solution:
Built a fully autonomous AI agent that:
• Extracts and enriches company data
• Scores lead relevance using GPT-4
• Pushes structured summaries to Slack and Airtable
• Triggers follow-up workflows based on score thresholds

Outcome:
• 90% reduction in manual screening workload
• Response time dropped from 2h to 5 minutes
• Conversion rate improved by 35%
• Agent now runs 24/7 without human supervision

Role: Lead architect and developer
Status: Live in production