It’s been a while since my last article about creating adaptable backends. In these past few months, I have been listening to and consuming as many courses and books as possible about modern GenAI. While the insights are often uncertain, today I decided to create this series to combine backend architecture with the world of inference.
GenAI has become one of the most impactful technologies of the last three years. I don’t remember any topic that has been on everyone’s lips for this long news, articles, movies, and all media in general talk endlessly about AI.
On the technical side, the noise is 3X louder. Technical professionals are often afraid of AI, while AI generalists struggle to apply it effectively in real-world projects.
The Engineer’s Guide to Ignoring AI Hype: Why I’m Writing This Series
In this series, I will focus on helping technical people who hold biases against AI and struggle to identify the knowledge transfer and parallelism that exists between AI and traditional software engineering.
I have met with AI specialists, data scientists, and professionals in related fields, and I see the same recurring patterns:
- They lack understanding of transferable skills: They sell AI frameworks (like LangChain or LlamaIndex) as a panacea of innovation, ignoring the reality that the framework fundamentals have existed for decades.
- If you don’t speak the language, you don’t know anything: This connects to the previous point. There is often a lack of understanding or loose communication if you don’t speak their jargon (Agents, RAG, etc.). This becomes an unfair reason to reject candidates in an interview simply for not knowing the latest trendy words.
- Struggling to adapt AI into projects: Teams are still struggling to integrate AI into production environments effectively.
- The MVP Trap: I have received offers in my inbox asking to bring AI expertise from a Backend and DevOps perspective because the AI specialist team created an MVP in 3 days, but now they need to scale it and don’t know how.
What is the differentiator of this series?
My goal is to un-overwhelm you. I will avoid explaining AI capabilities using obscure terms or biased marketing. We will analyze different tools and SaaS platforms, not to criticize them in an opinionated way, but to open your mind, bring awareness to the reality of the situation, and promote critical thinking.
The style will be similar to “The Adaptable Backend” series: practical, direct to the point, and filled with examples. Most importantly, we will create micro-projects that can be used in your private or enterprise products.
Ready to turn down the volume?
Next week, we stop being scared of the jargon and start looking at the code. We’re diving deep into Agent Frameworks, chunking strategies, and CI/CD pipelines for MLOps to see if they are a revolution or just a new wrapper for old patterns.
Subscribe to the [GenAI Series] and let’s un-overwhelm this industry together.
