A practical T.L. Draw conversation about selling SDKs as AI agents reshape developer tooling.
The Changelog uses T.L. Draw as a concrete case study in selling developer infrastructure while AI agents change how software gets built and discovered.

Changelog Media
500 episodes listed below
Founders and engineers on AI coding agents like Claude Code, and supply chain security.
The Changelog follows software development where it is actually being argued, shipped, funded, secured, and rethought. Its recent episodes move from TL Draw’s SDK business and Docker Hardened Images to AI agents, Python governance, open-source trust, and the changing job of writing code.
PRESS PLAY
The Changelog is a long-running interview podcast, part of the Changelog Media network alongside shows like JS Party, hosted by Adam and Jared, who conduct one-on-one conversations with software engineers, open source maintainers, and startup founders. Each episode centers on a single guest and a specific technical or business story: examples in this sample include InfluxDB co-founder Paul Dix describing a stalled attempt to debug AI-agent-written Rust code, TL Draw creator Steve Ruiz on running a small SDK company as agentic tools compress roadmaps, ESLint creator Nicholas Zakas on npm's supply-chain security failures, Docker's Tushar Jain on the release of free Hardened Images, and GitLab founder Sid Suri discussing his cancer treatment alongside his new venture KiloCode. A recurring throughline is the practical, day-to-day impact of AI coding agents on engineering workflows, code review, QA practices, and team psychology. Other recurring themes include open source sustainability, container and dependency supply-chain security, and the business models behind developer tools and SDKs. Episodes run long-form, typically 75 to 110 minutes, with a conversational structure and sponsor reads between segments. The show is aimed at practicing engineers, engineering leaders, and open source maintainers who want detailed, named accounts of how specific companies handle AI tooling adoption and security incidents rather than abstract commentary. Guests are typically founders, CTOs, or long-time maintainers with direct operational responsibility for the tools or incidents discussed, and occasionally an episode moves into a guest's personal story, as with Sid Suri's account of his medical treatment. Production is unscripted interview audio released on a roughly weekly cadence.
Made for: Practicing software engineers, engineering leaders, and open source maintainers tracking AI-assisted development and supply-chain security practices.
What sets it apart: Each episode is anchored to one named guest's specific incident or project — a debugged AI-agent failure, a security compromise, a product launch — rather than general tech-news commentary.
Software's best weekly news brief, deep technical interviews & talk show.
Based on 6 episodes we listened to · September 2026
Jared makes technical systems legible by tracing their engineering decisions, governance, business pressures, and operational risks.
The Changelog: Software Development, Open Source follows software outward from code into the institutions, hardware, and businesses that determine how it works in practice.
Wikipedia governance meets drone parachutes, with Jared asking how the machinery actually holds together.
Press play if you want
Short reviews from the PlayNext desk, based on the episodes we processed.
A practical T.L. Draw conversation about selling SDKs as AI agents reshape developer tooling.
The Changelog uses T.L. Draw as a concrete case study in selling developer infrastructure while AI agents change how software gets built and discovered.
Paul Dix turns agentic coding from slogan into a working engineer's QA problem.
Paul Dix brings a builder's patience to the messy middle of agentic software work: what happens after the demo, when coding agents are sent into real engineering sidequests and the results still need human judgment.
A candid engineering leadership conversation about accountability, team agency, and what changes when delivery outcomes land on your desk.
A loose, useful conversation about software leadership, centered on the shift from senior individual contributor influence to direct accountability for multiple teams.
A casual, developer-minded test of whether polished everyday apps are becoming natural born SaaS killers.
The Changelog turns a loose chat about personal subscriptions into a useful Technology conversation about what SaaS has to compete with now. The hosts compare the few paid tools they still keep, including a cheap shared grocery-list app, with H-E-B's polished shopping app, where…
A sharp, grounded conversation about why npm security feels overdue, and why leaving npm is harder than wanting to.
NPM security lands here as infrastructure work, not routine cleanup. The episode frames the registry as a piece of JavaScript's everyday plumbing, then asks why GitHub's response still feels slower and thinner than the problem deserves.
A practical look at Claude Code, agentic workflows, and why small software teams may need less SaaS than before.
The Changelog tracks software development as AI agents move from coding help toward broader computer work. Damien Tanner, founder of Pusher and now building LayerCode, gives the discussion a concrete center: Claude Code connected to browser access, LinkedIn, X, APIs, and a…
A loose, technical systems episode about slow builds, Fly instances, Varnish counters, and the awkward usefulness of AI-assisted debugging.
Kaizen! Let it crash is at its best when it treats operational mess as something worth poking at in public. The conversation opens with the familiar complaint that GitHub Actions can feel painfully slow, then moves into a live look at Fly instances, regional traffic, an SJC…
A funny, loose developer-tools hangout about slow CI, Grafana culture, and demos refusing to behave.
The GitHub problem starts with a very familiar irritation: GitHub Actions is everywhere, coding agents seem happy to generate workflows for it, and then the builds slow down just when feedback matters.

Long-form interviews on AI coding agents, infrastructure, and security, plus a recurring SED News roundup.

Stack Overflow's Ryan Donovan interviews founders and engineers on AI agents, protocols, and dev tools.

Michael Kennedy interviews Python maintainers on packaging, security, web tools, and AI.

Michael Kennedy and Brian Okken run down the week's Python tooling, typing, and packaging news.

Matt Stratton interviews DevOps practitioners on security, platform engineering, and trusting AI-written code.

Steve Ruiz, creator of TL Draw (a free whiteboard and canvas SDK), joins The Changelog to discuss running a software company as AI coding agents compress development timelines from quarters to days.
This episode of Changelog News covers Peter Steinberger joining OpenAI to advance AI agents, explores competing projects like ZeroClaw and Mimiclaw for low-cost AI hardware, and discusses Postgres tuning challenges and telnet security vulnerabilities.
GitHub Actions takes the first hit for being slow.
Paul Dix, co-founder of InfluxDB, joins The Changelog to discuss six months of using AI coding agents to develop InfluxDB3 and what he learned when things went wrong.
This episode of Changelog News covers recent tech stories, including AI.com's costly DDoS mishap, Mitchell Hashimoto's Vouch system for open-source trust, an AI agent team's expensive but flawed C compiler, a historical look at software development simplification, and critiques of AI-generated code.
The cold open is very Changelog.
This episode of Changelog News discusses the breaking of tech monoculture with new device diversity, covers security breaches like Notepad++, and explores topics such as outage transparency, forkable databases for AI, and developer productivity limits.
Personal SaaS gets small fast here.
npm feels neglected.
This episode of Changelog News covers tech developments including Clawdbot, an open-source AI assistant sparking excitement, the efficiency of Mac Minis for AI, insights on software engineering fundamentals, the end of curl's bug bounty, and advancements in Postgres extensions and package managers.
Home lab talk starts loose, then finds the good stuff.
Damien Tanner returns to The Changelog after almost 17 years.
This episode of ChangeLog News covers key software development topics, including AI agent addiction, social file systems, security risks in AI dependencies, Postgres best practices, and web dependency issues.
GitHub Actions feels painfully slow.
GitHub Actions is slow.
This episode covers tech news including Stack Overflow's decline due to AI, Linus Torvalds using AI for coding, AI agents enabling easier self-hosting, improved JSON formatting, and reflections on software nostalgia and design.
Sid Sabrandage is back on the Changelog, and this one has a lot packed into it.
This episode of Changelog News covers software development trends for 2026, including Brian Guthrie's Move Faster Manifesto with seven rules for speed, solutions to Cloud Code's context loss, Steve Yegge's Gastown multi-agent system, Depot's build acceleration insights, and how AI is making web development more creative and enjoyable.
Late December again.
Adam traces Ajay Kalkarni's path to Tiger Data.
Host Jared reviews the top tech highlights of 2025, selecting the coolest code, best prose, and favorite podcast episodes from each month.
Welcome to the Linux rabbit hole.
Zipline starts with medical supply delivery.
This episode explores the 'confident idiot' problem in AI, where models hallucinate and agree with users, and discusses solutions like the STEER SDK.
"Agents are the new developers." The cold open starts there, then gets into databases, generated code, and what infrastructure has to handle when agents fetch context, fan out work, start sandboxes, and run migrations.
Vogels looks at 2026 and beyond.
A developer argues that reducing ambiguity, not technical skills or experience, is what truly separates senior engineers from the rest.
Adam may be a distro hopper.
Wikipedia runs on trust.
A single line of Rust code brought down much of the internet when Cloudflare suffered a major outage last week.
Swarm is the word getting a closer look.
Spencer Chang makes the internet feel like something you can hold.
This episode of Changelog News covers multiple tech topics including the Zig programming language's innovative features, creativity techniques without drugs, debates about MCP servers versus Bash tools, BuildKit's Docker build optimization, and the value debate around AI-generated art.
ZSTD gets the Friends treatment.
Sean Gedeke's highlighted thread is about taste.
This episode of Changelog News covers emerging trends in software development, including the rise of Forward Deployed Engineers in AI, critiques of cloud platforms like AWS, and insights on LLM agents and React's dominance.
This episode discusses the performance issues with GitHub Actions in CI/CD workflows and introduces Namespace as a faster, drop-in alternative that caches dependencies and Docker layers to speed up builds and improve developer productivity.
Adam Jacob discusses how agentic systems are revolutionizing infrastructure management, shares insights on the AWS outage, debates the AI bubble, and reflects on personal lessons in open source.
Kaizen starts with tired complaints finally turning into control.