Ashpreet Bedi

Today I'll show you how to connect your agents to Slack.

Today I'm going to show you how to recursively improve your agents.

The future is here, just not evenly distributed.

FastAPI solved one of the biggest problems in software engineering: how to serve code as an API.

This week Thinking Machines released Inkling, an open-weights model built for customization and available for fine-tuning through its Tinker API.

In this article I'm going to show you how to build, manage and recursively improve your agents using coding agents.

The market is finally catching up to a simple fact, that every company needs to own its agent stack.

Agentic data labeling is one of those workhorse use-cases that keeps on delivering unexpected results.

Every team building agents ends up building an agent platform from scratch.

Coding agents have changed how we build software.

In 1973, Doug McIlroy added pipes to Unix.

Note: this post is about building your own agents (agentic software engineering), not about using coding agents.

Cloud platforms, data platforms, and now agent platforms.

For fifty years, software has been static.

The Key To Building Agentic Software That Works In the early 1940s, Bell Labs was building the national telephone network, the most complex technical system in the world at the time.

What is the simplest architecture for running a multi-agent system at scale?

Every company with 30+ people should have an internal data agent and today I'm making ours open-source: take Dash, run it in your cloud, and give your team access via Slack.

Every era of computing develops its own programming language.

Instead of a hello world tutorial, let me show you how to build a live multi-agent system.

Evals are the holy grail of AI engineering.

On Monday I introduced Learning Machines and yesterday I shared that it's finally working.

Every AI memory tool I've used is missing something.

It's almost 2026.

Here's a pattern I've been using to make my agents better without fine-tuning or retraining.

In this post, we’ll build a self-learning research agent that does something more useful than one-off web searches.

This post shows how to build a self-improving Text-to-SQL agent using dynamic context and "poor-man's continuous learning".

PSA: If you're serious about Agent Security, stop sending your transactional data to telemetry services.

Most people overcomplicate Agents.

✨ The intersection of software, systems and security engineering.

Andrej Karpathy shared on the Dwarkesh Podcast that LLMs don't have the equivalent of "culture".

✨ Lessons from 100s of conversations on AI products and how teams are adopting AI.

✨ The Multi-Agent Framework, Runtime, and UI.