🤖 Even AI can’t understand a customer the way a business analyst can

On December 18, we invite you to a meetup where we’ll delve into how AI actually affects cognitive processes and why the core skills of a business analyst are more valuable than ever.

🎤Firuza Ganieva – Lead Business Systems Analyst and Product Owner at Andersen
🎤Najaf Ganiev – Product Manager, UX Researcher, and POLIMI MBA 2025 graduate

🔗 Register here

⏰ Time: 19:30 (Baku time)/16:30 (СET)
⏳ Duration: 1.5 hours
🗣 Language: English
📍 Offline: Andersen’s office in Baku
💻 Online: The link to the stream will be sent to your email specified in the registration form

See you!

🌐 AI Landscape: What’s New for Business Analysts (December 2025)

AI hasn’t slowed down for the holidays. Over the last month, the main LLM platforms shipped updates that push us deeper into the era of agents, integrated workflows, and assistants with memory — exactly where BA/SA work lives every day.

Here’s a quick digest of what changed and why it matters 👇

🔹 Gemini 3 & Workspace Studio
Google rolled out Gemini 3 and Workspace Studio, bringing no-code AI agents directly into Gmail, Drive and Chat.
➡️ For BA/SA this means: you can treat “build an agent for this process” almost like “add a new form or macro” — for intake, triage, discovery briefs, and routine approvals.

🔹 Claude Opus 4.5
Anthropic’s new flagship model is tuned for deep reasoning, long documents, slides, spreadsheets and code.
➡️ For BA/SA: a strong “end-to-end” assistant that can help you move from stakeholder notes → structured requirements → API/data contracts → initial test scenarios.

🔹 Perplexity with Memory
Perplexity added assistants with Memory and access to multiple top models in one interface.
➡️ For BA/SA: a powerful research front-end where your project context persists across sessions — useful for ongoing market, regulation or competitor analysis.

🔹 Grok 4.1
xAI released Grok 4.1 with better reasoning and tight integration with X (Twitter) and live web content.
➡️ For BA/SA: a fast way to scan sentiment, reactions and early signals around products, policies or pricing moves.

What this means for BA/SA in practice:

1️⃣ Agents are becoming everyday tools, not just pilot projects.
2️⃣ Discovery and research become continuous threads, not one-off queries.
3️⃣ Documents, data and code now sit in a single loop.
4️⃣ The key skill shifts from “testing tools” to orchestrating workflows across several models.

💬 AI is evolving faster than our backlogs. The real advantage for BA/SA is not trying every new model, but embedding the right ones into daily processes.

How are you already using these updates in your projects?

#BusinessAnalysis #SystemAnalysis #AIforBA #AIAgents #LLM #Productivity #DigitalTransformation

🔥 TOP AI DEVELOPMENT EXPECTATIONS FOR 2026

AI is already part of IT delivery — but for BA and SA roles, the real question is what comes next.
Leading business and management media (The Economist, HBR, FT, Forbes) increasingly agree: 2026 will be about industrializing AI, not admiring it.
From that perspective, here are Top AI expectations that matter most for Business & System Analysts in IT:

1️⃣ From chatbots to agents
AI moves from “answering” to executing workflows.
➡️ BA/SA define permissions, boundaries, and failure scenarios.

2️⃣ AI embedded into workflows
Not another tool, but part of Jira, Confluence, QA, analytics.
➡️ Analysts design AI-augmented processes, not just prompts.

3️⃣ Governance becomes mandatory
Policies, audit trails, explainability, data boundaries.
➡️ BA/SA translate risk and regulation into system requirements.

4️⃣ RAG over generic intelligence
Grounded answers, trusted sources, traceability.
➡️ Analysts structure knowledge and define “source of truth.”

5️⃣ ROI pressure increases
Fewer pilots, more measurable outcomes.
➡️
BA/SA help decide where AI truly adds value — and where it doesn’t.

6️⃣ Human & societal impact matters
Trust, transparency, adoption, backlash risks.
➡️ Requirements increasingly include UX, accountability, and change management.

In 2026, BA and SA roles shift from “requirements writers” to AI-enabled system designers — shaping how AI actually works inside enterprise IT.

#AI2026 #BusinessAnalysis #SystemAnalysis #GenAI #AIGovernance #RAG #ITDelivery #DigitalTransformation

🟢 SURVIVORSHIP BIAS IN IT PROJECTS: LEARNING ONLY FROM WHAT “SURVIVED”

Survivorship bias is learning from success stories while missing everything that quietly failed.
In IT and product work this is everywhere — and it makes planning dangerously optimistic.

🔵 Typical BA case:
You’re asked to “reuse best practices” from past projects.
You open Confluence and find polished success cases: smooth rollouts, good clients, clean diagrams.
But you don’t see:
MVPs that were stopped
pilots that failed
systems that never scaled
projects that died mid-Discovery

So your new project copies winners
without seeing the invisible graveyard.
This leads to copied requirements that don’t fit context, and estimates that ignore failure risk.

✅ How to avoid survivorship bias:
Ask explicitly for failure or “stopped” project cases.
Review post-mortems, not just showcases.
Track rejected options and why they were rejected.
In retros, document failures as reusable knowledge.

Success is loud.
Failure is quiet — but often the best teacher.

#SurvivorshipBias #BusinessAnalysis #ITProjects #ProductManagement #ProjectManagement #LessonsLearned #RiskManagement #DecisionMaking #DeliveryReality #TechLeadership

🌐 Framing Effects in Dashboards: How Presentation Changes Decisions

Same data. Different framing. Different decision.
Framing effect means the way we present numbers changes how stakeholders interpret risk, priority, and value.

🟢 Typical BA case:
Two true statements about a release:
- “90% of users had no issues.”
- “10% of users experienced issues.”
One sounds safe. One sounds urgent.
Both are correct.
But the backlog will look very different depending on which you show.
Framing also happens through visuals:
- A chart with a tight Y-axis makes a minor change look like a crisis.
- A chart with a wide axis hides real issues.
- Stakeholders react to the frame, not the raw truth.

🟢 Typical SA case:
Compliance metrics shown as averages hide risky outliers.
The system looks healthy until a rare edge case causes an audit failure.

✅ How to avoid framing traps:
Show both sides when decisions are sensitive: success + risk.
Keep scales consistent across time.
Label charts clearly: what, who, when.
Add a one-line interpretation note:
- “This view emphasizes risk; paired chart shows stability.”
Ask yourself:
- “What decision could this framing push people toward?”

Dashboards are not neutral.
They are decision lenses.

#BI #DataViz #BA #SA #ProductManagement #AnalyticsTips