Experienced developer review: data loss, poor architecture, no accountability
I’m a retired federal government employee with 14 years of experience as a Microsoft Access application developer, and I evaluated Emergent AI hoping I could use AI for an idea I had to accelerate a basic app to use as a side hustle.
Following a worldwide outage, Emergent AI was unable to locate or restore an application I had built prior to the outage. The app was simply gone. There was no recovery option, no export, and no version history available. From a systems and data management perspective, this raises serious concerns about backend architecture, persistence, and basic safeguards that experienced developers would expect by default.
What was most concerning was the response. Emergent took no responsibility for the loss of my work. Instead of acknowledging the failure or offering remediation, they issued additional credits, which is meaningless when the underlying problem is trust in the platform’s ability to retain and manage data.
I repeatedly requested a refund and was told the following twice so this is obviously a canned response:
“I want to be transparent about how Emergent works. Each time you run a job, we pay for real computational power, and our usage-based model is based on that cost. For this reason, refunds are usually not possible as the credits are already heavily discounted.”
From a professional standpoint, this response avoids the core issue entirely. This was not a matter of compute usage—it was loss of user-created assets due to a platform failure. In regulated or government environments, this would be unacceptable.
I spent more time educating the AI Agent in the basics rather than building the same functionality manually. The interface and workflow felt immature and poorly thought out, especially for anyone with formal development experience. The scroll bar had major issues needing constant refreshing making messages from the AI agent difficult to read and respond to.
There is minimal transparency into what the AI is actually generating, which makes debugging and maintenance difficult or impossible. This is a serious limitation for anyone who has worked in regulated or enterprise environments.
Support and documentation are insufficient for professional users, let alone those who have no experience in app development. Questions were met with vague or muddled explanations rather than concrete guidance, which suggests the product is still experimental despite being marketed as a usable app builder.
Based on my background, I would not consider Emergent AI suitable for professional development, government work, or any environment that values data integrity and maintainability. In its current form, it feels more like a prototype than a serious development tool. I cannot recommend it.
14. januar 2026
Uopfordret anmeldelse