AI Builders Tell Their Stories

AI is accelerating the ability of builders to create and ship things. Here, we've collected stories from creators of trending AI projects as they look to build the future of software.

How AI Builders Approach Their Projects

These in-depth interviews with AI builders explain how they identified a key challenge in software development and designed projects to help engineering teams navigate the AI era.

Submit Your AI Builder Story

Are you building an open-source AI project for technical users? Share your story with us.

Submit Your Story Here

Article

MCP Memory Service

Builder Heinrich Krupp explains why his work as an enterprise data architect led him to create a project that turns agentic memory into infrastructure.

How to Think About Agentic Memory Job-to-Job

By default, AI agents still suffer from job-to-job “amnesia.” Open-source builder Heinrich Krupp explains how he tackled the problem.

Article

CodeMachine

Builder Moaz Muhammed discusses why he designed his project, CodeMachine, around agentic orchestration.

Why Orchestration May Be the Future of Agentic Development

The CodeMachine project for AI agent orchestration works on long-horizon tasks, but benefits from human product managers. Creator Moaz Muhammed explains.

Article

CoastyAI

AI builder Prateek Jannu's hypothesis is that computer use might be one of the most immediately enterprise-ready AI use cases. He explains his reasoning here.

Computer Use: AI’s Most Enterprise-Ready Use Case?

Enterprises are slow to adopt new technology. Founder Prateek Jannu explains why computer use might be the most immediately actionable enterprise AI use case.

Article

Skill Seekers

Developer Yusuf Karaaslan shares how a routine pass on his daily duties led him to design one of the fastest-growing repos on agentic skills.

Translating Data into AI Skills

Developer Yusuf Karaaslan explains how he ended up designing a system that translates a variety of data into AI skills.

Article

Noodles

One of the biggest bottlenecks in the age of cheaply spun-up AI code is understanding it. The founding team at Unslop.xyz discusses how their project uses visualization to decrypt reams of AI-generated code.

How Visualization Helps Unblock the AI Coding Bottleneck

The new bottleneck in software is understanding AI-generated code. The founders of Unslop.xyz, which created the open-source Noodles project, explain.

Article

HumanLayer

Founder Dexter Horthy digs into what's keeping developers from adopting agents as a completely mainstream component of their stack.

What's Missing to Make AI Agents Mainstream?

AI agents have the potential to revolutionize software development. HumanLayer founder Dexter Horthy explains what's missing.

Article

iii

Founder Mike Piccolo notes that back-end development needs to make the same shift front-end development did: Adopting a new core primitive.

How Back-End Engineering Is Evolving in the Age of AI

Is the future of back-end systems a streamlined, language-agnostic architecture to handle waves of agentic API calls? iii founder Mike Piccolo explains.

Article

WordLlama

Researcher-turned-engineer Lee Miller's theory is that key AI processes might be optimized to run on CPU rather than GPU. His WordLlama project uses NLP processing as a proof-of-concept.

Combining Context Engineering With Lightweight Compute

What happens when you combine fast natural language processing tools with a lightweight compute framework? WordLlama creator Lee Miller explains.

Article

TOON

The Token-Oriented Object Notation project was born when engineer Johann Schopplich identified an opportunity to reformat LLM inputs to save 30-60% of token costs.

How Data Serialization Improves AI Token Economics

TOON creator Johann Schopplich explains how better data formatting means better AI token economics, which isn’t the only thing for startup founders to consider.

Article

ROMA

Researcher Salah Alzu’bi discusses the role of task decomposition and orchestration in successful multi-agent frameworks like his team's project ROMA.

How to Approach Multi-Agent Frameworks

Multi-agent frameworks have potential for a variety of use cases. Researcher Salah Alzu’bi explains.

Want More AI Builder Stories?

Get more content for founders and builders straight to your inbox. Subscribe to Heavybit Updates.