AI-Powered Agile Leadership - Atelier [English]

Animated by Elena Vassilieva (Freelance) at Agile en Seine on october 7, 2025
This interactive workshop is designed to empower agile leaders with practical strategies and hands-on techniques for integrating AI into their leadership practices. The session will explore how AI agents can transform your role by providing data-driven coaching, automating routine tasks, and driving innovative solutions that enhance team performance and strategic decision-making.
Learning Objectives:
• Understand AI’s Impact: Learn the roles AI agents play in agile leadership, including coaching, automation, and innovation.
• Practical Application: Gain insights into real-world use cases that demonstrate measurable improvements in team efficiency and decision-making.
• Enhance Prompting Skills: Master advanced AI prompting techniques to generate actionable insights tailored to agile challenges.
• Implement Strategies: Develop actionable strategies that you can immediately apply to improve your leadership practices and drive better business outcomes.

1. 7 & 8 octobre2025 –Issy-les-Moulineaux
Leadership agile augmenté par l’IA
Elena, Vassilieva
Agile Leadership Trainer
AmaitusVR Studios

2. Merci à nos partenaires

3. AI Empowered Agile
Leadership
Technology is the enabler
Leadership is the driver

4. Elena Vassilieva
Agile Leadership for the real world.
Storytelling in the Metaverse
https://www.linkedin.com/in/elenavassilieva/
www.evassilieva.com
I help leaders to turn corporate
chaos into innovation
playgrounds

5. When you think of AI and leadership, what comes
to mind first: risk, opportunity, or both?
“Risk” -more concerned than excited
“Opportunity” -more excited than
concerned
“Both” -equally excited & concerned
(remainder)

6. When you think of AI and leadership, what
comes to mind first: risk, opportunity, or both?
Poll benchmarks (US, 2024–25, Pew Research Center)
Source: Pew Research Center, How the U.S. Public and AI Experts View Artificial Intelligence (surveys Aug–Oct 2024, published Apr 3, 2025).
38%
51%
11%
38%
15%
47%

7. Risks – Safety concerns about data privacy
In September, California signed the landmark SB-53 billmaking California the
first US state to mandate safety reporting for large AI developers, including
OpenAI, Anthropic, Meta, and Google DeepMind.
•The bill requires major AI firms toshare how they follow safety standards
and publish updates to their safety protocols within 30 days.
•It mandates thatcompanies report any incidents or crimes that have been
committed by an AI model without any human oversight or intervention,
such as cyberattacks and deceptive behavior.
•For business leaders, thismarks a move towards enforceable transparency
in one of the world’s largest AI markets.
•Companies operating in or selling to California will need to adjust their
compliance strategies quickly—potentially setting a precedent for other
states and countries.

8. Risks – Job loss due to the tasks’ automation
•In September Lufthansa announced to Cut 4,000 Jobs by 2030 Amid AI Push as the
airline continues to roll out artificial intelligence and automation of tasks.
•In September, Walmart CEO Issues Wake-Up Call: ‘AI Is Going to Change Literally
Every Job’ Head count expected to stay flat over next three years, despite growth
plans, as AI eliminates some roles and transforms others.
•In September, Accenture CEO Julie Sweet said the company will start “exiting”
employees who don’t adapt to AI.
•In August, Salesforce CEO Marc Benioff said the software company was able to cut
4,000 support roles thanks to AI.
•In June, Ford CEO Jim Farley predicted half of all white-collar workers in the US
would be replaced by AI.

9. Risk – lack of developers trust in AI work
Stackoverflow 2025 Dev Survey

10. AI impact on early-career workers
Since AI rollout, early-career
workers in highly AI exposed roles
(junior software developers and
customer service reps) saw ~13%
relative job declines.
The drop concentrated where AI
applies to automates tasks rather
than augments people knowledge.
https://digitaleconomy.stanford.edu/publications/canaries-in-the-coal-mine/

11. From AI Tools…To joint Human-AI Co-intelligence
What can we do to become leaders
that do not just automate tasks but
create Human- AI Co-intelligence to
transfer knowledge and expertise
and as a result to create space for
leaders to do their highest value
work?

12. Empowering AI with Human Expertise -
Meet my AI Partners
AI Agent - Elena’s AI twin AI Agent –
Scrum Master Companion
AI Agent - AI Copilot
and Sparring Partner

13. Let’s Ask My AI Partners
What is AI Empowered Agile Leadership?
AI Trigger* - a specific keyword, phrase, or condition that
causes the AI to respond in a particular way.
Follow-up Trigger* Triggers the AI to shorten a text.
Follow-up Reflection Prompt*
"How can AI amplify – not replace - my leadership?"

14. Ask AI Partner

15. Ask AI Partner

16. Ask AI Partner

17. Ask AI Partner

18. How can AI amplify Agile Leadership?

19. AI + Human Expertise = Fired Up Innovation
Why It Matters:
•True value emerges when human
expertise is transferred to AI models
by enabling joint human-AI co-
intelligence.
•Success in AI-driven competition
requires competence built on both
human and artificial intelligence.

20. AI + Human Expertise - Fired Up Innovation
The Shift:
•New type of AI-empowered leaders are essential
in driving innovation and guiding teams with AI
co-intelligence.
•The best use of AI: tackle problems we cannot
yet solve alone.
•To build stronger AI competencies, the
innovation and transformation experience of
Agile leaders must be transferred to AI Partners,
creating joint human–AI co-intelligence

21. Explicit knowledge vs. Tacit knowledge
Explicit
knowledge is
codified and easily
transferable.
Tacit knowledge is
experiential and
deeply rooted in
individuals.
Automate Augment

22. Adopt AI-first collaboration practices
Automate:
AI Tools take over → efficiency ↑, jobs ↓
Augment:
AI + people collaborate →productivity ↑,
skills ↑
More AI doesn’t always mean more
productivity.
A recent survey from BetterUp and Stanford
found that 40% of US workers reported
receiving AI-generated “workslop” over the
past month, which can require extra human
work to fix the issues.

23. AI that will work with you, not replace you
Define Goal & Use Case - Create an augmented workflow
1. Clarify: What do you want the AI agent to do? Document scope and limits (what it
should not do).
2. Design and curate Core Prompts (Knowledge Triggers)
- a small set of seed prompts that represent desired behaviors.
Each curated prompt should include:
Context (role, task, tone, audience).
Example input (what a user might ask).
Desired output (structured, accurate, aligned).
3. Run AI → Generate Response: Capture its response as-is.
4. Reflect & Compare: Compare AI response to your desired output.- AI Evals
5. Refine Prompts Add clarifiers or constraints to prompts.
6. Create Reflection Loop – AI agent self-correcting feedback loop.
7. AI Training log - “Best” prompts; Reflection rules; Templates for repeated tasks.

25. How AI Agents define AI
Empowered Agile
Leadership
Mindset of integrating AI tools
with human-centered, adaptive
leadership to enhance decision-
making, feedback loops,
collaboration, and aligning teams
around measurable business and
customer outcomes.

26. Five principles of AI-Empowered Agile Leadership
Augment, don’t replace. Accelerate outcomes. Keep humans at the
center. Experiment often. Build trust into every step
AI Empowered leaders guide people with tech, not for tech.

27. How to apply the five key principles to become a
confident Agile Leader in AI reality?
Risk Risk/Opportunity Opportunity
38%

28. AI as a Problem-Solver
Start with the business problem — not the model
Leaders define:
• The decision at hand
• Data required
• “Good enough” threshold
Measure success with:
•Faster decisions
•Fewer handoffs
•Clear progress on key metric

29. Generative Engineering Culture
Ship small, learn fast, scale what works.
Ship small, learn fast, scale what works
Leaders focus on:
•AI prompt design + workflows as products
•Bias checks + feedback loops
•Blameless retrospectives
Success signals: weekly experiments,
documented prompts, open feedback
culture

30. Diversity -
Diverse teams, data, and perspectives = better models +
fairer outcomes.
•What it means: Include varied roles,
backgrounds, and domain expertise in AI
implementation design and review.
•Leader’s focus: Mixed-discipline squads,
bias checks, user testing across segments.
•Success signals: Reduced error gaps across
user groups; broader adoption and
satisfaction.

31. Trust -
Trust is a feature—build it in, test it, and show your work.
•What it means: Human-in-the-loop for
consequential decisions. Trust is built by
overcommunication and feedback loops.
•Leader’s focus: Policies on disclosure,
escalation paths, safety checklists.
•Success signals: Regular feedback loop,
consistent standards, fewer escalations and
more experiments.

32. Using Your Product
Lead by example—be your product’s power user.
•What it means: Leaders and teams use the
same AI capabilities they ship to customers.
•Leader’s focus: Internal pilots, executive
usage dashboards, share “wins” and
“warts.”
•Success signals: Active internal usage,
shortened feedback loops, faster
product/fit improvements.

33. Using Your Own Products - Build empathy
through lived experience
Empathy and integrity
come from experiencing
your own product

34. AI doesn’t replace leadership. It
empowers it
•What this changes (Outcomes)
•Decision quality ↑ (better options, fewer
setbacks)
•Cycle time ↓ (from months/weeks to
days/hours)
•Empathy at scale (personalized comms, better
support)
•Culture of learning (experiments → evidence
→ adoption)
Agile-Empowered leaders aren't just adding AI tools to
their roadmap. They're rewriting it entirely.

35. Make AI part of the Team
•Cross-Functional Teams:AI and Human Co-Intelligence
•AI Literacy:Team members understand how to work
with AI, from prompts engineering to interpreting
outputs. Designing Reflections and Evals.
•Experimentation Culture:Encouraging teams to test,
fail, and iterate quickly to discover where AI adds the
most value.
•Focus:Build a culture for humans and AI work together.
That means redefining roles, creating shared goals
between humans and AI, and fostering a mindset of
continuous learning and adaptation.

36. Action Plan – Mindset shift
The mindset will define the next generation
of Agile leaders in an AI-powered world.
•Lead with confidence: augment people,
empower teams, and shape the AI future.
•AI doesn’t replace leaders—it amplifies
them.
•"The future of leadership is human + AI,
working better together.
ALIGN
ACTIVATE
AMPLIFY
ACCELERATE
GOVERN

37. Apply Five Principles in your own context to shift
from hesitation toward confidence
•Augment, don’t replace - - -> ALIGN
•Experiment often - - - -> ACTIVATE
•Keep humans at the center - - - ->AMPLIFY
•Accelerate outcomes - - - -> ACCELERATE
•Build trust into every step - - - -> GOVERN

38. Risk Implications for Managers
Align: Let AI draft the report, but I own
the insights.
Activate: Run a 2-week AI pilot and
share lessons.
Amplify: Use AI to handle routine tasks
so I spend more time coaching.
Accelerate: Speed up decision-making
with AI data summaries.
Govern: Set clear rules for when AI vs.
humans decide.

39. Action Plan:
Create and engage with AI Agent
1.Create three different persona/ one for each LLM
(ChatGPT, Claude/Sonet, Grok/DeepSeek
2.Train AI Agent on curated prompts/reflection
3.Design and implement Evals (defined data set with input
and expected output; code to evaluate AI responses or
use another LLM-as-Judge
4.Understand and evaluate traces (detailed records of
interaction between user and LLM)
5.Scale what works → sunset what doesn’t

40. Reflection Question:
What is your action plan to bring AI
partner working co-intelligently with the
team?
“AI doesn’t replace leadership. It empowers it.”

41. 7 & 8 octobre 2025 – Issy-les-Moulineaux