Не GPT єдиним: можливості AI в бізнес-аналізі | Вебінар з Тетяною Перловською

Штучний інтелект вже давно вийшов за межі простого генератора текстів. Його можливості активно проникають у сферу бізнес-аналізу – не як тренд, а як інструмент, що реально підсилює якість рішень, швидкість комунікації та точність роботи з вимогами.

На вебінарі ви отримаєте:
📌 Живу демонстрацію: як із фрагментарного голосового запиту від замовника (без структури й деталей) створити набір вимог за допомогою ШІ.
📌 Огляд найкорисніших інструментів AI, що дійсно допомагають у щоденній роботі
📌 Структурований підхід до використання ШІ у SDLC: що доречно автоматизувати, а де потрібна критична участь людини.
📌 Поради щодо впровадження AI у свою практику: з чого почати, як уникнути типових помилок і як підвищити продуктивність без втрати якості.


Спікерка: Тетяна Перловська
Business Analysis Manager @EPAM Systems
Engineering manager & EngX Productivity Expert

1. NOT ONLY GPT: AI CAPABILITIES IN
BUSINESS ANALYSIS

2. AGENDA
Why AI is a BA Game-Changer
Live demo
AI BA SDLC
What Machines Do vs. What Humans Do
2

3. ABOUT ME:
Tetiana Perlovska
Senior Business Analysis Manager
EngX AI Productivity Expert
17+ years in BA, 7+ years in Management
Proud owner of a maltipoo

4. WHY AI IS
A BA GAME-CHANGER

5. WHAT CONSUMES MOST OF YOUR TIME AS A BA?
5

6. AI ISN’T MAGIC—IT’S A
FORCE MULTIPLIER THAT
CUTS BUSYWORK BY 50%,
LETTING YOU FOCUS ON
STRATEGY

7. THREE PILLARS OF BA+AI
Automation of Routine Work
AI automates documentation and note-taking, allowing business analysts to
focus on strategic and high-value tasks.
Data Analysis
AI supports risk identification and recognizes historical patterns, enabling
informed decision-making through advanced data analysis.
Communication Enhancement
AI-powered visualization tools and structured information improve
collaboration and make communication more impactful and clear.
7

8. TURNING CHAOS
INTO CLARITY

9. Tool: Otter.ai, Chat GPT
Promt:
“Role: Senior Business Analyst
Task: Convert this stakeholder voice note into:
1. Clear problem statements
2. Functional requirements (numbered)
3. Implied needs
4. Data sources needed
Voice Note: [Insert above transcript] ”
Output:
**Problem Statement:**
Prevent overselling by automating low-inventory detection and alerts.
**Functional Requirements:**
1. **Real-time inventory tracking:** Monitor stock levels per SKU.
2. **Threshold-based alerts:** Notify managers when inventory < [X] units (configurable).
3. **Overselling block:** Restrict orders if stock ≤ 0.
4. **Dashboard visualization:**
- Live inventory heatmap (red = critical)
- Overselling incident history
**Implied Needs:**
- Customizable alert rules (email/SMS)
- Integration with sales channels (e-commerce, POS)
- Historical trend analysis ("Last quarter" suggests forecasting)
**Data Sources Needed:**
- Current inventory DB
- Sales transaction logs
- SKU master list
1.AI-Powered Note Structuring

10. 10
Output:
graph TD
A[New Order] --> B{Inventory Check}
B -->|Stock ≥ Order Qty| C[Process Payment]
B -->|Stock < Threshold| D[Trigger Manager Alert]
B -->|Stock = 0| E[Block Order & Notify Customer]
D --> F[Update Dashboard Alert Log]
C & E --> G[Update Inventory]
G --> H[Dashboard: Real-time Stock Levels]
2. Requirement → Visual Workflow
Tool: ChatGPT, Mermaid
Promt:
“Generate a diagram code (Mermaid.js) for:
1. Inventory check on order placement
2. Alert if stock < threshold
3. Block sale if stock = 0
4. Update dashboard ”

11. Output:
1. **As a** Warehouse Manager,
**I want** SMS/email alerts when inventory drops below 10 units per SKU,
**so that** I can replenish stock before shortages occur.
*Acceptance Criteria:*
- [ ] Alerts trigger within 30 seconds of threshold breach
- [ ] Configurable thresholds per SKU category
2. **As a** Sales Agent,
**I want** orders blocked automatically when stock = 0,
**so that** customers aren't sold unavailable items.
*Acceptance Criteria:*
- [ ] System shows "Out of Stock" message at checkout
- [ ] Backorder option offered if enabled
3. **As a** Operations Head,
**I want** a dashboard showing live inventory levels and shortage alerts,
**so that** I can prioritize resupply efforts.
*Acceptance Criteria:*
- [ ] Dashboard loads in < 2 seconds
- [ ] Color-coded risk zones (green/yellow/red)
3. Generate User Stories
Tool: ChatGPT
Prompt:
“Convert requirements into Agile user stories (INVEST
principle). Include acceptance criteria.
Key requirements:
- Low-stock alerts
- Overselling prevention
- Dashboard with real-time data”

12. AI BA SDLC

13. 13
SDLC Phase BA Activities GenAI Tools GenAI-Assisted Deliverables
1. Initiation / Discovery
Stakeholder analysis
Problem/opportunity definition
Business case creation
Elicitation planning
Risk analysis
Feasibility assessment
Use AI chatbots to simulate stakeholder interviews
ChatGPT, Claude, Notion AI, Microsoft
Copilot, Whimsical AI
Stakeholder maps & personas
Draft problem statements
Business case outlines
Risk registers
AI-generated SWOT/PESTLE Stakeholder
Q&A simulations
2. Requirements Elicitation
Interview/survey prep
Workshop planning
Eliciting business rules
Brainstorming sessions
Gathering use cases/user stories
Auto-transcribe & summarize sessions
ChatGPT, Fireflies.ai, Otter.ai, Dovetail, Scribe
AI, Miro AI
Elicitation guides
Meeting minutes and follow ups
First-draft user stories
Business rule catalogs
Workshop agendas & activity plans
3. Requirements Analysis
Categorizing & grouping requirements
Gap & impact analysis
Conflict resolution
Prioritization (MoSCoW, Kano, Value-Risk)
Traceability planning
Use GenAI to generate visual models and decision
logs
ChatGPT, LucidchartAI, Miro AI, NotionAI,
Claude
Requirements matrix
Gap/impact analysis tables
Conflict resolution logs
Traceability matrix drafts
Decision logs
Auto-generated process & data diagrams
4. Requirements Specification
- BRD/SRS writing- User story/scenario writing
(INVEST, BDD)- Defining glossary, assumptions,
constraints- Requirements validation with SMEs
Leverage AI to generate reusable templates
ChatGPT, Confluence AI, Word Copilot,
Grammarly, Notion AI
- BRD/SRS first drafts- User stories and
Gherkin syntax- Reusable specification
templates- Glossary/term lists- Validation
checklists

14. 14
SDLC Phase BA Activities GenAI Tools GenAI-Assisted Deliverables
5. Solution Design Support
Map requirements to design- Participate in
walkthroughs- Clarify business needs to
designers/developers- Review mockups/wireframes-
Define NFRs
Use AI to annotate wireframes and visualize
requirement mappings
ChatGPT, Figma AI, Miro AI, Mermaid Live
Editor
Requirement Design traceability-
Annotated UI/UX mockups- Draft NFRs
(performance, usability, etc.)- Visual flows
(Mermaid, UML)
6. Solution Evaluation & Testing
UAT support- Test case derivation- Defect triage-
Requirement validation- Regression impact analysis
Automate test mapping and generate test data
ChatGPT, TestRail AI Assist, Excel Copilot, Jira
Assist
UAT scripts- AI-generated test scenarios-
Requirement-test coverage reports-
Regression risk reports- AI-suggested test
data
7. Deployment & Support
Prepare transition documentation- Training &
onboarding- Release comms- Feedback collection-
Operational readiness check
Record training sessions and auto-generate help
content
ChatGPT, LoomAI, Otter.ai, SurveyMonkey
AI, NotionAI
Handover documentation- Training guides,
video scripts- AI-generated FAQs- Post-
deployment survey summaries- Knowledge
base articles
8. Continuous Improvement
Monitor KPIs- Analyze user feedback/behavior-
Groom backlog- Suggest enhancements- Lessons
learned capture
Use AI to auto-analyze feedback and recommend
next actions
Enable semantic search across project
documentation
ChatGPT, Power BI Copilot, Tableau GPT, Jira
Assist, Confluence AI, Notion AI
KPI dashboards with GenAI insights-
Groomed & rephrased backlog items-
Enhancement proposals- Semantic search
over project assets- Lessons learned
summaries- Auto-prioritized improvement
areas

15. BRING YOUR CASE

16. WHAT MACHINES DO
VS.
WHAT HUMANS DO

17. TASK AI BA
Data Gathering Automated collection Validation (relevance/accuracy)
Structuring Pattern recognition
Key term definition (business
glossary)
Visualization
Auto-generate
charts/diagrams
Stakeholder alignment (context
tuning)
Risk Identification Data-driven predictions
Contexte interprétation
(culture/politiques)
Solution Design Not applicable
Strategy + Negotiation (trade-
offs)
Responsibility Matrix

18. AI READS WORDS. THE ANALYST
READS BETWEEN THE LINES
18

19. IMPLEMENTATION: STRATEGY FOR YOUR TEAM
19
1. Feed AI Precisely
Bad: "Describe the system" → Vague, generic output
Good: "Write user stories for POS checkout: 'As a cashier, I
want to scan items faster with barcode batch processing...’”
Add constraints to prompts:
"Within PCI compliance standards," "Must support offline mode,"
"Max 3-second response time."
2. Start with Documentation
Use AI to auto-generate reports from raw data (Jira logs, user
interviews, telemetry):
"Analyze Sprint 23 bug data: Show top 3 modules by defect
density and suggest test coverage improvements."
Result: Free up 70%+ time for high-value analysis (e.g., root-
cause investigation of recurring issues).
3. Quality Control Guardrails
Adopt a Checklist:
Does this align with business goals?
Are acceptance criteria testable?
Is domain jargon consistent?
Schedule Pre-Meeting reviews: Dedicate 15 minutes to sanitize
AI outputs before sharing.
3 Pro Tips for start:

20. PITFALLS TO AVOID
20
Risk Consequence Mitigation Strategy
Over-reliance on AI"Floaty artifacts" (no substance)
Ground outputs: Always add "How does this
impact [KPI]?" to prompts.
Ignoring context Missed stakeholder emotions
Inject context: "Stakeholder is frustrated with
manual workarounds because..."
Skipping testing Hallucinated logic in production
Pilot first: Validate outputs on a low-risk "lab
project" (e.g., internal tool).

21. ТУТНЕМАЄ‘ЧАРІВНОЇКНОПКИ’ —
ЄСИСТЕМАСТВОРЕННЯСВОГОAI-
ПОМІЧНИКА, КОТРИЙПРАЦЮЄЯК
ВАШАДРУГАПАРАРУК
21

22. THANK YOU
22

23. ОНЛАЙНКУРС
2
3
Інтегровані AI-інструменти в повсякденні
робочі процеси
1
Швидкий й ефективний автоматизований
збір, аналіз і управління вимогами
Аналіз ринку, конкурентів та побудова
ефективних бізнес-кейсів за допомогою АІ
Унікальна знижка
10%за промокодом:
webE5
AI for Business Analysis
з 07.10.2025 до 28.10.2025
Діятиме до 05.09
Щовиотримаєтезавдякинавчанню: