# SynergyAI Dev Update 0x0002

By [SynergyAI](https://paragraph.com/@synergy-labs) · 2024-03-10

---

### **SynergyAI Dev Update 0x0002**

Welcome to the second development update of SynergyAI! It’s been quite a busy couple of weeks for the team, so let’s dive into what we’ve managed to ship 🛳️

> We’ve started onboarding users at scale  
> **Sign up now →** [**https://synergyai.io/waitlist**](https://synergyai.io/waitlist)
> 
> [https://synergyai.io/waitlist](https://synergyai.io/waitlist)

### **Summary 👀**

*   TON integration - now you can analyze tokens from The Open Network on SynergyAI
    
*   Migrated DB from MongoDB to Supabase
    
*   Reworked gamification
    
*   Adjusted API for running at scale
    
*   Resolved frontend bugs
    

### **Development Insights 💻**

**TON Integration**  
Since DeFi on TON is booming, the TON integration was brought up to the top of our agenda. That created a significant challenge for the development team, because the entire infrastructure was previously designed to be compatible with EVM blockchains.

There were two paths we considered taking:

1.  **Crappy and fast** - we would be able to launch in 2-3 days. The problem was that the codebase would quickly become unmaintainable and we would spend the next weeks digging ourselves out of the bad code architecture this way would require. The path required duplicating a lot of core components of SynergyAI and modifying them to support data from TON. While it would probably work, there was no way we would be able to sustain that in the long term.
    
2.  **Full multichain** - launch in 5-7 days, less headache down the road. This approach required adjusting the entire codebase of SynergyAI to not only work with all _EVM_ chains, but with virtually every type of on-chain data provided (after parsing). We decided it was the right way to do it, because it will allow us to integrate all new chains that the bull run will bring while having a stable core that won’t break the second we want to expand it.
    

A lot of SynergyAI’s features are still not live on TON - including portfolio tracking and live new pairs, but we’re working around the clock to deliver them as soon as possible. We prioritized launching fast and getting a well-working product with less features to the users rather than postponing the launch to integrate every single feature.

**Frontend Bugs**  
Users have reported a couple of minor frontend bugs to us. These included bad scaling of the top section on mobile devices, search modal not being always fully visible and tokens not displaying correctly in OpenGraph image. All of these should be fixed now.

**Automated Testing**  
We have set up automated tests for most of the UI elements using Playwright. This change will allow us to encounter less bugs in the future. Our current implementation runs all the tests on each git commit, making it virtually impossible to crash something when committing new changes.

**Reliability Engineering**  
As SynergyAI started growing and more users starting sending requests to our backend, it was clear that our implementation wasn’t fast and reliable enough.  
  
The first problem was the server. Up until this point all of SynergyAI was hosted on multiple servers (some cloud-based, others local). This solution was not fully reliable because of the relatively low uptime of local servers.  
  
We moved the entire infrastructure to a cloud server provided by [Hetzner](https://www.hetzner.com/).

**API Rework**  
Since the volume of requests sent to our backend largely increased, we also had to optimize our software implementation. These changes include moving our databases from MongoDB to Supabase (mainly for data subscriptions), developing a new system of worker (indexer) to API querying and the implementation of the full TradingView interface.

**Gamification Fixes**  
We were informed by some of our beta testers that they encountered problems with the gamification system (where XP was not translated to levels and some quests were not showing completion).  
  
These problems have all been resolved with a thorough rework of the gamification system which was a direct result of the migration from MongoDB to Supabase.

### **Socials 📱**

Thank you for reading this update! Come back next week to hear more about the development of SynergyAI.

[https://twitter.com/synergyai\_io](https://twitter.com/synergyai_io)

[https://t.me/safeguard\_synergyai](https://t.me/safeguard_synergyai)

[https://synergyai.io/](https://synergyai.io/)

> Karol Proksa, 10 Mar 2024

---

*Originally published on [SynergyAI](https://paragraph.com/@synergy-labs/synergyai-dev-update-0x0002)*