Building an In-House Lovable (engineering.merciyanis.com)

In a significant shift for software development, a company has transformed its engineering workflow by leveraging advanced AI models like Anthropic's Opus 4.5 and OpenAI's GPT-5.2. Previously reliant on manual coding, engineers now utilize a fleet of AI agents that can autonomously execute complex tasks, drastically increasing productivity. The results of their experiment revealed that the vast majority of tasks could be completed with minimal human intervention, achieving quality levels suitable for production at an unprecedented speed—up to 32 times faster than junior developers for simpler tasks. This newfound efficiency has allowed a 25-person team to handle work that would have typically required hundreds of engineers. This transformation is significant for the AI/ML community as it showcases the practical application of cutting-edge models in real-world scenarios, challenging traditional roles within engineering. The authors emphasize that while coding has become more streamlined, the essence of engineering—problem-solving and conceptualizing solutions—remains crucial. With the rise of platform engineering, the focus is now on maintaining code quality and coherence at scale, requiring engineers to adapt and evolve their skill sets. This shift underscores the balance between human ingenuity and AI capabilities, prompting discussions on redefining roles and the overall organizational structure in tech companies.