Episode 11 · September 15, 2026 · 52:09
How AI Saved 69 Weeks of Work: How AI Turns Experts Into Builders
Guest: Gabor Kis, Business Systems Analyst in Change Management at DTCC
About this episode
What if the hardest part of AI at work is not the model, but getting a clear prompt and a repeatable process before you touch the fancy tools? Gabor Kis, Business Systems Analyst in Change Management at DTCC, joins Samantha Viola Wilson and Brittany George to walk how a longtime Xerox data architect reskilled into AI, won an internal hackathon on AWS Kiro, and now teaches business builders to treat English as their coding language. He shares prompting as the number-one skill, beginner-to-agentic stages (Aha, Mirage, AI Vampire, Synergy), the Grafana story that helped reclaim roughly 69 weeks, and why specs beat vibe coding for real change work.
Chapters
- Welcome to Intelligence Resources with Samantha Viola Wilson and Brittany George (0:00)
- Brittany welcomes Gabor through daughter Grace (0:11)
- From Xerox data architect to DTCC change management (0:39)
- Google AI Studio interview prep and landing the role (1:37)
- Betting the hackathon on AWS Kiro (2:05)
- DTCC scale and why change management cannot fail (2:55)
- Winning the AI hackathon and starting to teach (3:52)
- English as the coding language for business builders (5:29)
- The aha moment and the Grafana project named Grace (5:55)
- How 45 seconds times 69 spreadsheets saved ~69 weeks (7:22)
- Hesitant colleagues and the AI adoption gradient (8:05)
- What is the most important AI skill? (8:48)
- Samantha and Brittany on prompting, process, and patience (9:00)
- Prompting as communication, and why the tool matters less at first (10:26)
- Cheat sheet: make the model write a 10-out-of-10 prompt (11:24)
- Probabilistic answers, not perfect deterministic ones (13:02)
- Coursera tracks and the Vanderbilt prompting course (13:56)
- Stages: Aha, Mirage, AI Vampire, Synergy (16:00)
- Nine-agent software factory and Sonny the orchestrator (17:41)
- OpenClaw experiment, 8.5M tokens, and pulling the plug (19:13)
- What is AWS Kiro, and why enterprise stacks pick Amazon and Microsoft (21:25)
- Spec development versus vibe coding (23:14)
- Free credits, pricing tiers, and small-business cost math (24:37)
- Copilot aversion, and using Copilot as a prompt workshop (27:09)
- Change management: build it, they will come, teach teachers (29:58)
- CAB GPT hackathon deep dive: risk and blast radius (32:20)
- AI security fears for PHI, SSNs, and financial data (34:52)
- Post-quantum computing and harvest-now decrypt-later (36:17)
- What quantum computing means (the maze analogy) (38:17)
- When quantum and AI meet (40:04)
- Human in the loop, presence, and empathy (42:31)
- Always-on work, motherhood, and tech resentment (45:05)
- New stage: slave to the machine (46:51)
- Personal AI use: attachment style prompt (49:51)
- Close, LinkedIn CTA, and Tuesday goodbye (51:06)
Connect with Gabor: https://www.linkedin.com/in/gaborkis
#PromptEngineering #AIForBusiness #ChangeManagement #ArtificialIntelligence #IntelligenceResources #AWS #Kiro #FutureOfWork #ChatGPT