Diversity: Two steps forward, one step back - Group talk (Panel Outline)

A Craft 2025 diversity-and-inclusion panel, moderated by Gabriella Machkolár (executive with 20+ years in financial services, digital transformation, and AI-driven innovation; passionate about D&I and STEM early education). Panelists: Noémi (graduate/junior product manager at Tesco, relatively new to tech), Arina (product manager at EPAM, Budapest-based, planning to open a Women in Big Data chapter), and Zach Pendleton (chief architect at Instructure, education technology). Title theme: progress isn’t a straight line. Format: a topic-threaded conversation interleaved with Slido audience interaction (a word cloud, a poll) and two audience pledges. This outline threads the conversation by topic.

1. Introductions and Why Diversity Matters

1.1 The Slido word-cloud prompt

• Gabriella opens by asking the audience via Slido: “What do you think about diversity and inclusion right now — the first word that comes to mind?” (noting it may differ from before).

1.2 Arina — a safe place to speak

• Most important: create an atmosphere of being safe — a safe place to speak and be heard.
• At and outside EPAM she speaks about diversity styles of leadership (regardless of gender) and cross-cultural communication.
• “There is no drama” — you should feel safe to speak about it.

1.3 Zach — diverse teams perform best

• Over a career managing teams that grew from just him to 20, 30, even 100 people, the most diverse teams performed best — best ideas, widest range of opinions and life experiences.
• Diversity is moral and drives business outcomes and healthier organizations.

1.4 Noémi — always something to learn

• Different perspectives and backgrounds mean “there’s always new things we can learn from each other.”
• The more diverse the team, the broader its views.

1.5 The title’s meaning

• Gabriella: progress is “not always a straight line” — not a neat progression from stage to stage.
• Arina: much has been done, but the further we go the more challenges appear; some initiatives have failed, and we need an honest look to acknowledge problems and talk about them.

2. AI as Both Risk and Help

2.1 AI in recruitment can hurt

• Arina: in the age of AI (used across domains, including recruitment), ask whether AI boosts diversity or the reverse.
• Example: Amazon halted an HR system because it downgraded résumés containing the word “women’s.”
• Be attentive and think critically about these tools.

2.2 The EU AI Act

• Zach commends the European Union’s AI Act for calling out human-resource use cases as high-risk and requiring additional scrutiny.
• Don’t let AI’s promise become “another step back” — keep humans in the loop in important processes.

2.3 “AI is what you teach it”

• Noémi: AI can pick up exclusion patterns — “AI is what you teach it,” so it isn’t always objective. “Just as much as it could help, it could hurt.”

2.4 The moderator’s two AI concerns

• Gabriella notes AI panels at conferences are often all-male.
• And she read that 94% of the administrative workforce is done by women — the first jobs to go with AI’s advance. What can companies do?

3. Widening the Pipeline

3.1 Broaden who you target

• Noémi (relaying Zach’s pre-panel point): if you keep targeting the same people, you get the same numbers — change the perspective and broaden who and how you target.

3.2 Provide learning at work, not just “on your own time”

• Zach: it’s easy for an org to just tell employees “learn the skill outside work.”
• But candidates from diverse backgrounds often face challenges that don’t map onto “traditional white men with spouse/extended-family support for child-rearing and elder care.”
• Serious investment means providing learning experiences at work, since such candidates may not have extra time outside work.

3.3 Lead by example

• Arina: showcase good examples — if you represent a minority group, don’t be afraid to take a stand, mentor, sponsor.
• “Sometimes we think we’re alone… but it isn’t like that” — communicate with peers and leadership.
• AI could even analyze senior leaders’ career paths to find common patterns and experiment with more diverse backgrounds.

3.4 Representation lowers barriers

• Noémi: lack of representation makes people see a barrier — reasons not to join, upskill, or learn.

3.5 Promote from within

• Zach: hiring senior roles only externally means “inheriting every other company’s bad decisions” about diversity (interviewing only stereotypical candidates who already had that job).
• Fix: invest heavily in teaching and learning so employees can grow and be promoted from within — “I don’t see another way to fix this.”
• Gabriella and Arina: “there should be a funnel.”

4. First Audience Pledge

4.1 The pledge

• Gabriella asks the audience to promise:
  • Women: within the coming months, ask within your organization for at least one opportunity to learn/talk about AI, or start an AI chapter (inside or outside the firm).
  • Men: be an ally and invite women into activities.
• Nearly 100% of hands went up — “a wonderful start.”

5. Reading the Word Cloud

5.1 Gabriella — innovation and color

• Picks “innovation”: diversity (cultural, gender, background, any differentiation) brings color, different opinions and perspectives — good for innovation.

5.2 Arina — “skipped” people

• Struck by “many different people” and “skipped”: in a corporation, imagine how many different people, perspectives, and potential innovations are skipped every day — and how scary that only a certain group decides “what is good, what is bad.”
• Examples of male-only design: cars that can’t be opened with long nails (only male engineers, untested for women); crash-test dummies modeled on males, so women and children get more injuries.

5.3 Zach — accessibility as diversity

• Glad to see accessibility counted as part of D&I.
• Software engineers shouldn’t split the world into “abled” and a small “disabled” group — “we are all temporarily able-bodied”; investing in those who need help is “an investment in our future selves.”

5.4 Zach — cancelled DEI programs

• On the US pushback and cancellations: one good thing is clarity — “you now know which organizations never cared about DEI.”
• They had programs “because they thought that’s what they needed to hire talent, not because they cared.”
• Remaining challenge: bring those companies along and get them genuinely invested.

5.5 Noémi — programs need real support

• Having a program isn’t enough — the company must enable change: if a participant proposes an inclusive innovation, leadership can’t just say “no, we can’t do that.”

5.6 Noémi — the “pressure” word

• Pressure: underrepresented groups feel more pressure to prove themselves.
• Example from a Women’s Day talk: a junior PM, a woman of color and a foreigner, saw no one like her in leadership and felt she had to “put in 10,000%” to reach that level.

6. Loneliness in Male-Dominated Rooms

6.1 Noémi — the observation

• New to tech, initially overwhelmed; a few months in, she looked around and “couldn’t name one woman I’d worked with in three months.”
• Not discrimination — a very inclusive group — but “it would be nice to have more of us here.”
• It’s not only about being a woman: a foreigner, a person of color, anyone can feel they “don’t have their little minority group to support” them.
• Since then, Tesco put a lot of effort into hiring women — now she sees women in the office, “which makes me happy” — the representation feels nice.

6.2 Arina — loneliness intensifies at senior levels

• At junior levels there’s hope of gender balance; higher up there are fewer minority people, and ”that’s the real loneliness” — competition is higher and stronger, with no sponsorship.
• Two-way fix:
  • Share your feelings: “I’m a bit lost, there’s a boys’ chit-chat I don’t understand — help me be part of the team informally, not just formally.”
  • Teammates should recognize the pain of that loneliness and be the first to extend a hand: “How do you feel today? Let’s have a coffee, let’s discuss work issues over coffee.” It must be a dialogue, not one-way.

6.3 Zach — hire senior roles from within (callback)

• Reinforces the pipeline point: only hiring senior people externally inherits other companies’ bad diversity decisions.
• Invest in teaching/learning so people can be promoted from within.

7. Practices to Start Early and Widen the Pool

7.1 Tesco’s “Girls’ Day”

• Noémi: Tesco ran a Girls’ Day bringing high-school girls in to show what tech work is like — broadening horizons before careers even start.

7.2 Part-time for every role

• Every Tesco position is now offered as part-time too — flex time around life stage, having children, caring for someone ill, or being ill yourself — widening the hiring pool beyond gender/ethnicity.

7.3 Programs emphasizing human connection

• Zach: successful programs focus less on formal training and more on space to ask questions, feel safe, and build connections that benefit people throughout their time.

7.4 Coaching a blended leadership style

• Arina: a vice president who advocates for diversity encourages his pool to be proactive and unafraid, and even teaches female directors “how to interrupt.”
• Women are often perceived as “nice girls smiling,” where interrupting/assertiveness seems bad.
• He inspires a blended leadership style — sometimes kind, sometimes pushy — “otherwise you will not reach your goals, and that’s okay.”

8. The KPMG “Hard Work” Finding

8.1 The finding

• Gabriella: a newly published KPMG study found that for women in the workplace, “hard work” has overtaken leadership/team-management skills as the most important thing — suggesting it’s getting harder for women to thrive.

8.2 Noémi — why “hard work”?

• Finds it “really interesting” — what does “just hard work” even mean, and why would it be most important? It signals things are changing and getting more difficult for women.

8.3 Zach — “hard work” is an unmeasurable, biased metric

• Organizations feel increased pressure to show profit and deliver value.
• He’s nervous about metrics like “hard work” because it’s hard to measure — a manager tends to see “hard workers” as people who work like me, look like me, speak like me, unintentionally leaving people out.
• Better: get concrete and measurable about performance criteria.

8.4 Noémi — merit-based, not KPI-based

• Diversity has become a metric, but be careful: you can’t exclude the majority just to hit metrics.
• Hiring must be merit-based — find the genuine good fit; increasing the pool lets you find amazing minority candidates.
• Don’t “overdo it”; keep standards stable so women aren’t criticized “for being a woman at the workplace.”

8.5 Arina — smart work and unequal home load

• We fulfill many roles (employee/business owner, mom, spouse, friend) — sometimes you should “work smart, not hard.”
• Strategic KPIs may need a hard push, but that sacrifices other life spheres — and women are usually the first to sacrifice, being the first/only to handle kids, household, elder care.
• There’s no equity in the division of home responsibilities; “it’s not about corporate life, it’s about the whole life.”

8.6 Noémi — hire for perspective, not KPIs

• D&I groups can provide support and adjust for minorities’ lives.
• “KPIs are not everything” — hire diverse groups for the perspective, not to hit diversity KPIs; an all-white-male company has “one restricted perspective” and misses cultures and life experiences.

9. The Audience Poll: Main Limitations

9.1 The poll

• Gabriella runs a Slido poll on the main limitations regarding D&I (five options).

9.2 “No women/minorities interested in tech” — reframed

• The most frequent answer was “no women/minorities interested in tech positions.”
• Noémi reframes: the real question is “are you targeting the right women, or enough women?”
• Zach: “If you hire where you’ve always hired, you get the candidates you’ve always had” — find people where they are, often with non-traditional education/experience backgrounds.

9.3 Education overcomes self-doubt

• Arina: minorities may say “I’m not a good fit for tech, it’s not my field,” but proper education and managing biases “can be a win.”

9.4 Inclusive-coworkers result

• Zach is happy many report inclusive coworkers — but “not everyone” does; “those of you who don’t — my company’s hiring, come see me.”

9.5 Early orientation and toys

• Gabriella: it goes back to early orientation — girls given “girly toys,” boys given Legos.
• Let girls try coding early; she’s seen girls try it for an hour and say “I can do that and I love it.”
• Also good: women changing jobs, upskilling/reskilling into technology.

10. Best Advice to Minorities

10.1 Zach — widen the pipeline and solicit voices

• Same themes: widen your pipeline; include more people in conversations even within the company.
• Leaders fall into “communicating the way we always have,” which biases toward those comfortable speaking up — actively solicit different opinions.

10.2 Noémi — be brave, be the role model

• “If you don’t feel represented, you can be the change and be a role model to all who come after you. Just be brave.”

10.3 Arina — you are not alone

• “You are not alone” (repeated for emphasis). Many groups struggle; you should feel safe because you’re not alone.
• Balance: don’t stay silent thinking your case is unique, but don’t “shout too loudly and be too annoying” either — find help, seek support, speak with leadership, and promote collaboration across teams.

10.4 Gabriella — seek mentoring and associations

• Minorities should seek mentoring and associations/groups of minorities who can show the path, provide support, and share advice — “it’s good to see people who have already made it.”

11. Cross-Cultural / Geographically Distributed Teams

11.1 Noémi — divided teams need effort

• Her team has development in Hungary and product management in the UK; it “feels very divided.”
• Constant effort is needed to be “a team together,” not “the Hungarians and the British people.”

11.2 Arina — individualist vs. collectivist cultures, ground rules

• International teams may mix individualist and collectivist members.
• Speak openly about differences (“we come from different cultures, but we need common ground”) and work out ground rules that vary team to team.

11.3 Zach — dedicated time, never all-but-one co-located

• Make dedicated time for cross-geography communication — it builds relationships and trust.
• What “works horribly”: everyone in one location except one person — “a great way to drive people out of your organization.”
• Invest in understanding cultures.

11.4 The Indian-food anecdote

• Zach’s worst example: a British coworker and an Indian coworker had a huge fight because the British coworker insisted Indian food was a British invention — “so culturally insensitive.” “Don’t do that.”

11.5 Managing conflicts without bad intent

• Zach: when there’s no bad intent, pause and move the conversation up a level — talk about what was intended and understood, not what was literally said. Understanding where the breakdown was (often cultural) lets you navigate around it.
• Noémi: meeting in person (even once or twice) really helps build connections.

12. Neurodiversity

12.1 Noémi — set expectations, discuss accommodations

• Clearly set expectations and communicate as the neurodivergent person; discuss role goals and possible accommodations with your manager.

12.2 Arina — understand strengths and fit

• Communication is key to accommodate; understand a person’s strengths and where they’ll be the perfect fit.

12.3 Zach — educate, then accommodate

• First understand the whole topic — run educational sessions on what neurodiversity is, its types, and how to fit people into the workplace.
• Make accommodations — e.g., people with ADHD dislike noise, so create a quiet office corner (like the “silent part of the train”) where they can work comfortably.

13. Closing

13.1 Biggest challenges in Hungary

• Gabriella: essentially all of the above — be more inclusive.

13.2 The pledge reminder

• Don’t forget your pledges: within one month, reach out to and include other diversity colleagues. “Just one action would start the changes.”

13.3 Thanks

• Gabriella thanks panelists Arina, Noémi, and Zach.

People & References Cited

• Gabriella Machkolár — moderator; executive (financial services, digital transformation, AI-driven innovation); D&I and STEM early-education advocate.
• Noémi — graduate/junior product manager at Tesco.
• Arina — product manager at EPAM (Budapest); planning a Women in Big Data chapter.
• Zach Pendleton — chief architect at Instructure (education technology).
• Companies / orgs: Tesco (Girls’ Day, part-time roles, women-hiring push), EPAM, Instructure, Amazon (halted biased HR/résumé tool), KPMG (study: “hard work” now tops leadership skills for women), European Union (AI Act), Women in Big Data.
• Tools / platforms: Slido (word cloud and poll), AI/LLMs in recruitment.
• Concepts: safe space to speak, diverse teams drive business outcomes, AI bias (“AI is what you teach it”), EU AI Act flagging HR as high-risk / humans-in-the-loop, widening the pipeline, learning at work vs. on your own time, promote-from-within vs. inheriting others’ bad decisions, representation reducing barriers, loneliness (especially at senior levels), pressure to prove (“10,000%”), cancelled DEI revealing who never cared, accessibility as diversity / “temporarily able-bodied,” early orientation and toys (girly toys vs. Legos), blended/assertive leadership style (“how to interrupt”), merit-based hiring vs. diversity KPIs, hire for perspective, unequal home/care burden, individualist vs. collectivist cultures, team ground rules, moving conflict “up a level” to intent, in-person meetings, neurodiversity accommodations (quiet ADHD corner), the ally/opportunity pledges.

Video: https://www.youtube.com/watch?v=hFVXZfTzlsw — Transcript via yt-transcript.sh; outline generated from the transcript.