Bringing the JavaScript community together through free resources, events, and trainings! | Visit @ThisDotLabs for JavaScript consulting & help | [email protected]
This Dot Labs has two sides.
One is our Tech Consultancy side, where we have our highly technical teams working with clients, ensuring they never miss a deadline again!
The other side is our media side, where we have podcasts, a YouTube channel filled with workshops, and more!
. @RobOcell will be joining the AI Leadership Exchange on June 12th in Chamblee, GA.
As VP of Innovation at @ThisDotLabs, Rob helps organizations navigate the intersection of AI and software development, with a focus on engineering workflows, developer tooling, and technology strategy.
Join us as engineering leaders discuss what it takes to successfully adopt and scale AI across modern organizations.
CTOs, CIOs, and VPs of Engineering can request an invite at [email protected].
Not all #AI coding agents work the same way, even when they're powered by the same model.
A coding harness is the layer that manages context, exposes tools, executes actions, and coordinates work between the model and its environment. It's what allows an #LLM to interact with files, terminals, browsers, and other systems.
As organizations evaluate tools like @Claudeai Code, #Codex, #Pi, and others, the harness is becoming an increasingly important part of the conversation.
More on our blog: https://t.co/JBTh3HSstq
As #AI-assisted workflows become more complex, teams are starting to run into new challenges around skill composition, reusable sub-skills, context management, and long-term workflow reliability.
Thereβs still a lot of experimentation happening around how best to structure these systems, and evals are increasingly becoming the mechanism teams use to determine what actually performs well in practice.
If your team is building with coding agents or AI workflow orchestration, check out this discussion with @MGechev and @RobOcell:
Not all AI coding agents work the same way, even when they're powered by the same model.
The biggest differences often come from the harness: how it manages context, handles tool calls, requests approvals, and interacts with your codebase.
As teams explore tools like @Claudeai Code, #Codex, @OpenCode, and others, those implementation details are becoming harder to ignore.
One of the biggest advantages AI brings to engineering workflows happens before implementation even starts.
Instead of beginning with a blank page, teams can use agents to generate directions, recommendations, and early prototypes that accelerate ideation and decision-making across product and engineering.
Check out the full State of AI livestream recording featuring Mike Ryan, @DKundel, @ElliottFouts, and @ladyleet for more:
Christmas in June is back at Modern Web ATL on June 11 in Alpharetta! Our favorite Summer tradition!
Hear @BrandonMathis share why Pi has become his preferred coding agent and how he's customizing it with GPT-5.5. Learn how Kiah Tolliver from @localstack uses chaos engineering to build more resilient systems through intentional failure testing.
+ A fireside chat with @JonFontanez and @RobOcell on #AI engineering, developer tooling, and the future of software development.
Plus cookies, holiday outfits, and great conversations with the Atlanta developer community. ππͺ
Link in the replies!
Tracking #AI usage doesnβt automatically mean teams are getting value from AI.
Join @BrandonMathis to look at reports around Amazon tracking token usage across engineering teams and compares it to the old practice of measuring lines of code written.
The bigger question: are organizations measuring meaningful outcomes, or just measuring activity?
Engineers are learning how to orchestrate parallel systems of work instead of individually implementing every task themselves.
The teams seeing the biggest gains are rethinking workflows, delegation, validation, and how engineering collaborates with product and design.
Check out the full State of AI livestream recording featuring Mike Ryan, @DKundel, @ElliottFouts, and @ladyleet for more π
One of the biggest advantages #AI brings to engineering workflows happens before implementation even starts.
Instead of beginning with a blank page, teams can use agents to generate directions, recommendations, and early prototypes that accelerate ideation and decision-making across product and engineering.
Check out the full State of AI livestream recording featuring Mike Ryan, @DKundel, @ElliottFouts, and @ladyleet for more:
Many #AI-assisted development workflows can produce impressive results initially, but maintaining reliability, efficiency, and consistency becomes much harder as teams scale them across real engineering environments.
As organizations operationalize AI, conversations are increasingly shifting toward evals, token efficiency, context management, workflow reliability, and how to structure skills so agents behave more consistently over time.
Check out the full episode of the Modern Web Podcast featuring @MGechev and @RobOcell:
#AI-assisted development workflows are introducing a new set of tradeoffs and operational risks as teams increasingly adopt Markdown-based skills, MCP-connected systems, and more agentic tooling.
As these workflows become more embedded in real engineering environments, questions around reliability, security boundaries, workflow entropy, validation, third-party integrations, and long-term maintainability are becoming significantly more important. Harnesses, governance layers, and structured orchestration are also emerging as critical components for keeping AI-assisted systems predictable, auditable, and operationally safe at scale.
The shift is changing not just how teams build software, but how they think about oversight, trust, and operational control in AI-powered development environments.
Full episode here: https://t.co/h4YIXsh4f3
#AI-assisted development workflows are introducing a new set of tradeoffs and operational risks as teams increasingly adopt Markdown-based skills, MCP-connected systems, and more agentic tooling.
As these workflows become more embedded in real engineering environments, questions around reliability, security boundaries, workflow entropy, validation, third-party integrations, and long-term maintainability are becoming significantly more important. Harnesses, governance layers, and structured orchestration are also emerging as critical components for keeping AI-assisted systems predictable, auditable, and operationally safe at scale.
The shift is changing not just how teams build software, but how they think about oversight, trust, and operational control in AI-powered development environments.