Atlassian vs Linear

HR Trends in Fast-Growing Startups: AI, Skills, Transparency

A Leapsome report analyzing 100 high-growth tech companies identifies four HR trends. First, HR is becoming the next AI investment area, with 81% of companies expecting AI skills in HR staff and new roles like Atlassian's Chief People and AI Enablement Officer. Second, HR teams are not shrinking but using AI to increase productivity while maintaining small, dense teams. Third, AI expertise increases salaries by an average of 56%, and salary transparency is becoming standard, with 74% publishing salary info and companies like Duolingo and Klarna listing bands. Finally, HR leadership is moving into the C-suite, with 56% having a VP-level HR head, and some CPOs coming from non-HR backgrounds. The report highlights a focus on strategic HR work and mission centrality.

• Leapsome's report analyzed 100 high-growth tech companies and 5,000 data points.
• 81% of the companies expect AI knowledge from HR staff.
• Employees with AI skills earn on average 56% more than comparable professionals.

Ticketpane Launches Persistent VS Code Jira Client

Geary Works, a one-person company in New Mexico, has launched Ticketpane, a Jira Cloud client for VS Code, Cursor, and Windsurf. Unlike the official Atlassian extension which users report disconnects frequently, Ticketpane focuses on staying signed in by storing API tokens in the editor's encrypted secret storage. The free tier offers basic issue viewing, while Pro features like status changes and issue creation cost a one-time $19 per developer. The product is independent and not affiliated with Atlassian.

• Ticketpane is a Jira Cloud client for VS Code, Cursor, and Windsurf, launched by Geary Works.
• Ticketpane uses API tokens stored in encrypted secret storage to maintain sign-in, avoiding OAuth.
• The official Atlassian VS Code extension has a 2.55-star rating, with top complaints about disconnects.

10-K Financial Filing Analysis for Atlassian (2026-08-14)

Die Atlassian Corporation hat ihren Form 10-K-Bericht für das am 30. Juni 2026 endende Geschäftsjahr vorgelegt und dabei den strategischen Übergang zu einer AI-First-Cloud-Plattform im Rahmen der Initiative 'Atlassian Ascend' hervorgehoben. Dieser Plan sieht die schrittweise Einstellung der Data Center-Produktlinie vor: Der Neuverkauf wurde im März 2026 eingestellt, Produkterweiterungen enden im März 2028 und der vollständige Support läuft im März 2029 aus, um Kunden zur Cloud-Migration zu bewegen. Der Bericht dokumentiert zudem einen tiefgreifenden Wechsel in der Führungsebene (Ausscheiden von President, CFO und CTO) sowie die Autorisierung eines Aktienrückkaufprogramms im Umfang von 2,5 Mrd. USD.

• Im Rahmen der 'Atlassian Ascend'-Initiative wurden Neuverkäufe von Data Center-Produkten im März 2026 eingestellt; Erweiterungen enden im März 2028 und der vollständige Support läuft im März 2029 aus.
• Die Führungsriege wurde umfassend umstrukturiert: Der bisherige President, CFO und CTO schieden aus und wurden durch einen neuen Chief Product and AI Officer sowie einen neuen CFO ersetzt.
• Das Board autorisierte im Oktober 2025 ein neues Aktienrückkaufprogramm über 2,5 Mrd. USD, das im März 2026 startete.

AI Research Roundup: Terminal Agents, Cloudflare Traffic, Nanochat

This newsletter edition covers recent AI research and tools. Key items include a paper on terminal agent training with self-improving tasks, a method for compressing agent screen memory, a tool for generating editable 3D scenes, and a model that predicts environment responses. Cloudflare's analysis of 206 million web sessions reveals mixed human-agent control, impacting bot detection. A new C file implementation runs a 744B parameter model efficiently, and a code graph tool supports 150+ languages. Additionally, Karpathy's nanochat project trains a GPT-2-class model for $48, and a benchmark shows Apple's SpeechAnalyzer outperforming Whisper Small. The newsletter also highlights videos on model serving and an internal agent at Linear.

• Cloudflare recorded 206 million Precursor evaluations in a day, showing sessions can shift between human and automated control.
• Colibri runs a 744B parameter MoE model from a single C file, using VRAM, RAM, and NVMe as tiers.
• Karpathy's nanochat trains a GPT-2-class model for about $48 on eight H100s.

Coding sessions: Linear Agent can now set up, run, and test your code

Linear Agent can now set up, run, and test your code before returning its work. That means fewer handoffs and changes that are further along when they come back to you.