Experience
Overview
Enterprise workflow, financial management, and CRM platform built with React 19, TypeScript, Material UI, Django 5.0, Django REST Framework, and AWS SQS. Featuring live QuickBooks Online synchronization, Google Drive file storage, CRM pipeline management, employee shifts tracking, and multi-module reports and analytics dashboards.
Responsibilities
Led the end-to-end full-stack development of the SaaS application. Designed the frontend state management (Redux Toolkit, React Query) and scalable RESTful API endpoints. Configured AWS SQS queue polling, Change Data Capture (CDC) syncing, third-party integrations, and dynamic role-based access control (RBAC).
Challenges
- Ensuring real-time data consistency and high-velocity synchronization with QuickBooks Online while staying within API rate limits and preventing request timeouts.
- Securing sensitive customer credentials and managing automated token refresh cycles for both QuickBooks and Google Drive OAuth flows.
- Handling complex, fuzzy search queries across large, high-concurrency database tables (inventory, contacts, transactions) efficiently.
- Offloading time-consuming data processing tasks from the main HTTP request thread to maintain a responsive user experience.
- Designing a reliable, stateful clock-in/out console (Kiosk) to manage employee time tracking and shift durations securely.
- Calculating complex, multi-module business intelligence metrics (utilization rates, conversion funnels, cash flows) in real-time without introducing database locks or loading lags.
Solutions
- Implemented QuickBooks Change Data Capture (CDC) sync, checking updates since the last recorded timestamp and caching sync state in a QBSync model.
- Utilized symmetric database-level cryptography (django-fernet-fields) to store access and refresh tokens, alongside automated background refresh checks.
- Optimized query performance by annotating PostgreSQL TrigramSimilarity for fast fuzzy string matching on names, SKUs, and descriptions, with SQLite icontains fallbacks.
- Designed a dedicated queue runner polling AWS SQS to execute background sync tasks and handle webhook events asynchronously.
- Implemented a PIN-authorized Kiosk dashboard with session locking (/employee/kiosk/login) to manage employee shifts and clock-in/out time logs.
- Engineered a database-level caching mechanism (AnalyticsCache model) with a 5-minute TTL to store pre-calculated metrics, completely bypassing external API call latency during page loads.
Impact
- Successfully synchronized 10 critical accounting entities (Invoices, Bills, Inventory Items, Payments, etc.) between Allyvia and QuickBooks Online, ensuring ledger consistency.
- Decoupled heavy processing tasks into SQS-backed background jobs, keeping web server latency under 200ms.
- Created an automated onboarding system that lowered user setup times by sending temporary credentials and tracking active status.
- Encrypted and secured third-party OAuth access tokens across all active company profiles, preventing credential leakage.
- Delivered dynamic menu filtering, route guarding, and UI action permissions based on subscription tiers and user roles.
- Provisioned a secure physical kiosk station allowing zero-friction clock-in/out time entries and generating automated shift durations.
- Enabled full client pipeline management (leads, deals, and notes) linked directly to company documents and financial invoices.
- Delivered sub-200ms response times for all reports pages (Finance, Inventory, Employee, CRM, Overview) through efficient backend caching and database aggregates.