Work — Rexallar

FinTech

Real-time payments dashboard

A legacy batch-based reconciliation system was replaced with a real-time dashboard processing 2M+ transactions daily, giving finance teams instant visibility into settlement status.

Challenge: Daily batch jobs delayed reconciliation by up to 12 hours.

Solution: Event-driven architecture with streaming data pipelines.

Result: Reconciliation time dropped from 12 hours to under 2 minutes.

HealthTech

Patient scheduling platform

Designed and built a HIPAA-compliant scheduling application now used by 40+ clinics to manage appointments, reminders, and intake forms.

Challenge: High no-show rates and manual phone-based scheduling.

Solution: Self-serve booking with automated SMS/email reminders.

Result: No-show rate reduced by 28% within the first quarter.

E-commerce

Headless storefront migration

Migrated a monolithic storefront onto a headless commerce architecture, decoupling frontend and backend for faster iteration.

Challenge: Slow page loads and long release cycles on the legacy platform.

Solution: Headless CMS + edge-rendered storefront.

Result: Page speed improved 3x; conversion rate up 14%.

Logistics

Fleet tracking & route optimization

Built a real-time fleet tracking system with route optimization, cutting fuel costs and improving delivery time estimates.

Challenge: Manual dispatch and inaccurate delivery windows.

Solution: Live GPS tracking with an optimization engine for routing.

Result: Fuel costs down 17%, on-time delivery up 22%.

SaaS

Multi-tenant analytics platform

Built a B2B analytics SaaS from scratch, including billing, role-based access, and embeddable dashboards for customers.

Challenge: No existing product — needed a market-ready MVP in 4 months.

Solution: Multi-tenant architecture with Stripe billing and white-label dashboards.

Result: Shipped on time; onboarded first 15 paying customers pre-launch.

FinTech

AI-assisted underwriting tool

Integrated an LLM-powered risk assessment assistant into an existing loan origination workflow.

Challenge: Manual document review slowed down loan approvals.

Solution: RAG-based document analysis with human-in-the-loop review.

Result: Average review time cut from 3 days to 4 hours.