TAG Weekly Roundup #4

Editor's Note

Atlassian's latest research reveals 98% of executives worry their teams aren't effectively using AI to eliminate silos. Meanwhile…their "AI teammates" are delivering measurable results: 

• 2,000 questions answered by onboarding agents in month one

• 80% of dev work auto populated with PR Generator 

Also this week, the emergence of the "Agent Engineer" role, a career opportunity for that small percentage of Technical BAs with coding and ML experience.  

📻News

Atlassian Proves AI Teammates Deliver Real Results

Atlassian’s latest report shows their AI agents are solving genuine workplace friction. As mentioned, their onboarding agent handled 2,000 questions in its first month, whilst PR Generator auto-completes 80% of dev work. in the room. The report also states that 2.4 billion hours are wasted searching for information each year within Fortune 500 companies.

56% of workers say the only way to get information is asking someone or scheduling a meeting!! Read more

🧰Tools

Eraser DiagramGPT - AI That Actually Understands Business Processes

Generate technical diagrams in seconds from plain English or code - flowcharts, entity relationship diagrams, cloud architecture diagrams and more. Built by the Eraser team using OpenAI's GPT-4.

Free tier available, though professional features require subscription. Worth testing for requirements documentation workflows.

The Personal MBA - Business Analysis Fundamentals

Not a tool, more a reference read.  

Josh Kaufman's accessible guide breaks down essential business concepts typically taught in MBA programmes. I’m currently at negotiation.  

📽️Video

LangChain CEO Discusses the “Agent Engineer” Role

Harrison Chase argues we're witnessing the emergence of a new profession - The Agent Engineer.  The role combines four critical competencies: 

1. Prompting. The ability to interact with LLMs effectively

2. Traditional Engineering. Building reliable systems and data pipelines (skills traditionally found in software engineers and data engineers)

3. Product.  Domain knowledge to understand user workflows that need to be automated and put into an AI agent

4. Machine Learning. Skills traditionally found in data scientists and researchers

For that small cohort of Technical BA’s with coding and ML experience, this is an alternate opportunity over the well-trodden product / project routes. However, for the majority, it's  a new start and a steep learning spiral!

72 Hours of AI Agent Workshops - Distilled for BA’s

This overview from Tina Hang reveals a systematic approach for automating business workflows with AI agents. The Business Workflow for Automation framework (15:55) presents a methodology for identifying and implementing successful automation:

The Five-Step Framework:

1. Observe - Look at what people are doing in the business process

2. Decompose - Break down workflows into smaller manageable tasks

3. Map - Identify task relationships and draw out branches and interactions

4. Prototype - Build an initial simple version of the workflow in an agentic way

5. Evaluate and Iterate - Write evals to identify where agents aren't behaving as expected, then improve through prompt changes, models, or tools

The key insight - domain knowledge becomes the competitive advantage that makes agents genuinely useful rather than technically impressive.

Did you Know? Senior BAs Command Premium Salaries: Business Analysts with 7+ years of experience earn $100,809 compared to $70,001 for those with less than one year—demonstrating clear career progression value. Source: Built in Salary Analysis, 2025

Till next week.