Undated
Tamar Yehoshua says Atlassian's tests showed that coding agents using its teamwork graph produced 44% better quality and used 48% fewer tokens than using MCP alone.
“Increased by forty four, and very importantly, the token usage went down by forty eight. Because you don't have to give as much context to the model”
Tamar Yehoshua · How Snap Built a Production System Where Agents Write 90% of Code, 0:08:31
Undated
Tamar Yehoshua says Atlassian's teamwork graph contains over 200 billion entities with links between them.
“our teamwork graph has over 200 billion, uh, entities in it and there are links between it.”
Tamar Yehoshua · How Snap Built a Production System Where Agents Write 90% of Code, 0:11:42
Undated
Tamar Yehoshua says 70% of projects built during Atlassian's quarterly AI Builders Week continue to be used afterward.
“70% of what's built during Builders Week actually gets used later.”
Tamar Yehoshua · How Snap Built a Production System Where Agents Write 90% of Code, 0:46:48
Undated
Tamar Yehoshua says that before AI, roughly 20% of software development time was spent coding, with planning before and review and operations after.
“about approximately twenty of your time was spent coding and you had left of code and you have right of code.”
Tamar Yehoshua · How Snap Built a Production System Where Agents Write 90% of Code, 0:36:04
Undated
Tamar Yehoshua says she replaced monthly OKR (objectives and key results) meetings with an agent that reviews Confluence, Jira and Slack and reports which goals are green or red.
“I, for example, no longer have monthly OKR meetings. I have an agent that runs that gives me the updated OKRs”
Tamar Yehoshua · How Snap Built a Production System Where Agents Write 90% of Code, 0:42:58
Undated
Atlassian's MCP and command-line interfaces expose over 500 tools to AI agents, according to Tamar Yehoshua.
“So, like our MCP and CLIs have over 500 tools from Atlassian that we expose to the agents,”
Tamar Yehoshua · How Snap Built a Production System Where Agents Write 90% of Code, 0:04:14
Undated
Tamar Yehoshua says about 95% of what can be done in Atlassian's user interface can also be done through MCP or CLI.
“We've gotten to the point where 95% of everything you can do in the UI, you can do through MCP or SCLI so that you can access that.”
Tamar Yehoshua · Atlassian Chief AI Officer: Building Products for Agents and Humans, 0:01:57
Undated
Yehoshua's team at Atlassian also runs corporate development and Atlassian Ventures.
“my team also runs, uh, Corp Dev and Atlassian Ventures.”
Tamar Yehoshua · Atlassian Chief AI Officer: Building Products for Agents and Humans, 0:28:46
Undated
Yehoshua says that before AI, roughly 20% of the software development lifecycle was spent coding.
“Pre AI, uh, if you think of the the software development life cycle, about approximately twenty of your time was spent coding”
Tamar Yehoshua · Atlassian Chief AI Officer: Building Products for Agents and Humans, 0:36:04
Undated
Yehoshua says 70% of projects built during Atlassian's AI Builders Week continue to be used afterward.
“70% of what's built during Builders Week actually gets used later.”
Tamar Yehoshua · Atlassian Chief AI Officer: Building Products for Agents and Humans, 0:46:48