ProjectsSignalStack
01 / Project Details
LivePersonal analytics / dashboard platform
SignalStack
SignalStack is a full-stack personal analytics platform that unifies GitHub, Google Analytics, TikTok, YouTube, and market data into one polished, product-grade dashboard — connect once, monitor everything.
RoleFull-Stack Engineer
Duration5 Weeks
ClientPersonal Project
Year2025
Project Summary
A full-stack analytics product that brings GitHub, Google Analytics, TikTok, YouTube, and CoinGecko signals into one customizable dashboard with queue-backed syncs, encrypted token handling, and production-ready observability.
02 / Snapshot
Connected source types
5 integrations
Normalized metric taxonomy
13 metric keys
Sync automation
Hourly scheduled jobs + on-demand provider syncs
Core product surfaces
Dashboard, Integrations, Jobs, Preferences, Onboarding
03 / Overview
Building a personal analytics experience that feels cohesive, trustworthy, and operational.
SignalStack was shaped as an end-to-end product for people who want a clearer view of their digital footprint — without juggling five dashboards to get it. Connect your data sources, sync them, inspect trends, and shape the view around what matters.
Challenge
The underlying problem was fragmentation. GitHub activity lives in one ecosystem, GA4 traffic in another, social growth in separate vendor dashboards, and external reference data somewhere else entirely. Even when the data exists, it is not organized into a unified narrative. The challenge was to design a product that could bring these signals together without becoming cluttered, brittle, or operationally opaque.
Solution
I approached the product as a unified signal layer rather than a collection of isolated integrations. Each provider feeds a normalized metrics model, background sync jobs handle ingestion, and the frontend translates that data into a focused, customizable dashboard. Just as importantly, the system exposes its own operational state through jobs, health checks, sync status, and integration-specific configuration so the product is both usable and maintainable.
04 / Stack & Scope
What shipped
Tech Stack
Next.jsTypeScriptNestJSPrismaPostgreSQLTimescaleDBRedisBullMQRechartsTailwind CSSSentry
Team Context
Solo Full-Stack Engineer
Deliverables
- Responsive landing page with a premium dark visual style
- Authenticated product shell with onboarding, dashboard, integrations, jobs, and preferences flows
- OAuth-based connection flows for GitHub, Google Analytics, TikTok, and YouTube
- Zero-OAuth enablement path for CoinGecko
- Integration-specific setup controls such as GA4 property selection, GitHub repo scoping, and YouTube channel selection
- Customizable dashboard with date-range filtering, integration filtering, metric filtering, hidden-chart control, and chart reordering
- Background sync infrastructure using BullMQ workers and scheduled internal sync triggers
- Job health and run-history monitoring with worker heartbeat, queue status, durations, attempts, and error visibility
- Secure token handling with AES-256-GCM encryption before persistence
- Production deployment topology spanning Vercel, Render, Neon, Redis, and CI-gated deploy automation
Key Highlights
- Metric points from very different providers are stored in a consistent schema, making cross-source visualization and future expansion much easier.
- The product does not hide background complexity. Users can inspect sync runs, job statuses, and worker health in a dedicated Jobs surface.
- The integration flow is not one-size-fits-all. GitHub supports repo targeting, Google Analytics supports property selection, and YouTube supports active channel selection.
- Preferences and layout controls allow the product to adapt to the user instead of forcing a fixed reporting experience.
05 / Process
From product framing to final polish.
Step 01
Define the signal model
The first step was framing the product around one core outcome: helping a user understand personal performance across multiple ecosystems from a single surface. That required defining which signals mattered, how they should be normalized, and how the dashboard should feel focused instead of overloaded.
Step 02
Design connection and setup flows
Once the sources were clear, the next challenge was integration UX. Each provider needed a setup path that felt intentional: OAuth where appropriate, lightweight enablement where possible, and source-specific configuration for properties, repositories, and channels. This kept the onboarding experience practical instead of generic.
Step 03
Build the sync and storage pipeline
The backend was structured around a queue-backed sync workflow. The API enqueues jobs, the worker processes provider-specific sync logic, and normalized metric points are written into Postgres with TimescaleDB-friendly time-series access patterns. Token handling, refresh logic, encrypted storage, and job status updates were treated as first-class concerns.
Step 04
Turn infrastructure into product experience
The final step was translating backend capability into product confidence. That meant not only rendering charts, but also adding useful controls, surfacing sync status, exposing worker heartbeat, supporting preferences, and tightening the UI so the application feels deliberate from landing page to authenticated dashboard.
06 / Gallery
Selected product views
07 / Closing Notes
This project focused on turning fragmented personal data into a single product system — one that feels clean in the UI, disciplined in the architecture, and credible in real-world use.
Product Thinking
The experience is framed around a clear user job: understanding personal performance across multiple platforms from one place.
Engineering Depth
The stack includes OAuth handling, encrypted token storage, job queues, worker health tracking, time-series storage, and scheduled sync automation.
Extensibility
The provider-based architecture and normalized metric model make the product easier to evolve into a broader analytics platform over time.
08 / Next Project