There is a lot of talk these days of peril for the software engineering industry. The fundamental thesis is that you will soon be able to dispatch work to an agent, or a ““swarm”” of agents to build your software.
In practice, agents need to be managed just like humans. Except they move 100x faster, and break 100x more in the process. They cannot be held accountable, be invested, or feel passion.
I’m far more interested in how we can leverage LLMs and agentic tools to exponentially improve the output and impact of our human colleagues. Here I will share the tools and methodologies we use to power our 100+ person engineering team. And its not just the engineers: our entire organization leverages AI to collaborate deeply and build together.
Everything I share can be readily constructed with simple off-the-shelf tools, not tied to any one company or model. It all shares one central thesis: invest in your engineers, give them first class tools designed to reduce friction, enrich context, and inspire at every turn.
When this technology is applied well, you can strike a precise balance and achieve unmatched speed with a high quality bar.
AI doesn’t have to be a boogeyman.
Key takeaways:
-
Inspiration that software engineering has a bright future as a field.
-
Real workflows they can take back and apply to their own projects to enrich them.
-
An understanding of the nuance of working with AI with helpful dos and don’ts.
Target audience:
- Anyone building software, especially those on larger teams or at large organizations.