Hi everyone, I’m Tariq Mehmood. I’m currently in my first semester as a computer science student, and for the past few months, I’ve been building my first major production project: a Finnish electricity price-tracking website Sähkötänään.
In my university classes, we are diving deep into C++, memory management, and core algorithms. But applying those foundational concepts to a live web application with real-time data fetching has been an entirely different beast.
I wanted to share a bit about the architecture hurdles I’m facing and see how more experienced devs handle the transition from "student projects" to scalable tools.
The Project
Sähkötänään is designed to help Finnish consumers track fluctuating exchange electricity (spot) prices. The core feature is a live data dashboard that pulls hourly pricing via API and feeds that data into interactive calculators (e.g., calculating the exact cost of running a sauna or home heating based on the current hour's rate).
The Architecture Hurdle: Managing Live Data
Initially, I thought building a dashboard just meant making a fetch() call and rendering a chart. The reality hit me hard when
I started dealing with:
- Rate Limits & Caching: I can't hammer the regional energy APIs every time a user loads the page. I'm currently working through the best ways to cache that hourly data on my backend and serve it efficiently to the frontend.
- State Management in Calculators: The calculators need to instantly react to both user inputs (like their specific contract margin) and the live API data.
- The AI Assist: As the logic got heavier, I started integrating AI tools (like Codex) to help scaffold the core data-handling logic while I focused on the overall architecture.
The Pivot: From Solo Dev to Technical Manager
The biggest lesson I've learned recently isn't even about code. It’s about scope.
To make the site an authoritative resource, it needs content—articles on optimizing home heating, contract comparisons, and daily price analyses. I quickly realized I couldn't write hundreds of articles, run a forum outreach campaign, and build a scalable backend at the same time.
I recently brought on a small team of three writers and an outreach specialist. Writing Standard Operating Procedures (SOPs) for them to ensure quality without relying on spammy AI content was my first real taste of being a technical founder. Delegating the content means I can finally step back and focus entirely on the codebase.
Questions for the Community
As I continue to build out the backend to support this new influx of traffic and content, I’d love to hear from those who have built similar data-heavy dashboards:
- Caching Strategies: What is your go-to architecture for caching and serving third-party hourly API data without over-engineering it?
- AI Integration: For those using AI to assist with coding, do you prefer to have the AI write your core API logic, or do you handle the data yourself and use AI strictly for frontend components?
- Scaling: Any advice for a student on making sure a newly launched dashboard doesn't immediately crash when traffic spikes?
Top comments (0)