Arnab Bose

For 18 years, @asana has helped teams coordinate complex work. AI Teammates build on that foundation, operating inside real workflows with the same plan, context, and goals as the teams they support.

We spent months working directly with customers to understand how AI Teammates should collaborate across an entire team, build on past decisions, and operate within clear guardrails.

The result: 30+ prebuilt AI Teammates and 160+ Skills for specific roles and industries, ready to work from day one.

There are lot of ways to record meetings but only one end to end solution that infers follow up items and automatically connects them to your projects, portfolios and human-agent teams: @asana AI Meetings

The primary bottleneck in software development has shifted from writing code to alignment, verification, and clear intent - I cover @asana's pov and why we built Command in my interview with @MTSlive here: youtube.com/watch?v=BZV_-l…

Excited to take the stage soon at @asana Work Innovation Summit New York! I’ll be talking about how @stackai + Asana bring together complex multi system orchestration and organizational context — and how with every completed workflow we compound knowledge.

@asana Command + @OpenAI GPT-6 Astra helped @stackai engineering cut a browser agent's cost 76x and made it 5x faster by keeping its cache intact.

GPT-6 Astra , working in Codex, did most of the work: it audited the code, instrumented every request, ran quick tests to identify the variables that mattered, refactored the code to run workflows in parallel, launched the runs and analyzed the traces. Humans set the goal and the standards and reviewed the conclusions. Command, Asana's software delivery platform, was the system of record: every session's requests, traces and results were recorded there, so we could analyze the full study afterward, and the findings became tickets, pull requests and reviewed changes that shipped.