--- title: "Reconstructed Slides: Moving away from Agile: What's Next – Martin Harrysson & Natasha Maniar, McKinsey & Company" category: "slides" video_id: "SZStlIhyTCY" sourceLabels: ["Cropped public YouTube video frames", "Local OpenCV slide-region detection", "Local RapidOCR"] --- # Reconstructed Slides: Moving away from Agile: What's Next – Martin Harrysson & Natasha Maniar, McKinsey & Company ## Source Video [Moving away from Agile: What's Next – Martin Harrysson & Natasha Maniar, McKinsey & Company](https://www.youtube.com/watch?v=SZStlIhyTCY) ## Method This deck is reconstructed from the existing video frame captures by detecting likely slide regions with OpenCV, cropping/upscaling those regions, deduplicating similar crops, and OCRing the cropped slide images locally. It is a cleaner companion to the full-stage frame deck. ## Reconstructed Slides ![[assets/reconstructed-slides/SZStlIhyTCY/slide-002.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-002.html) - AI slide classifier: `content_slide` confidence `0.97` - Text source: advanced OCR `rapidocr-live/bright-screen/contrast`. - OCR decision: ready — Dense text-heavy slide with chart labels and small supporting images. Slide text: > New technologies have given rise to new software dev methodologies > Pre-2000s 2000s 2010s 2020s > Tech breakthrough PCs Mainframes, Web, client-server APls,mobile Cloud, Al coding assistants > methodologies Software dev Waterfall Agile dev platform dev Product and Al-native dev > CEEE > MciOnsey&Comgany > Coaesummit ![[assets/reconstructed-slides/SZStlIhyTCY/slide-003.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-003.html) - AI slide classifier: `content_slide` confidence `0.99` - Text source: advanced OCR `rapidocr-live/bright-screen/contrast`. - OCR decision: ready — Full text-heavy slide with multiple labeled sections and images. Slide text: > New technologies have given rise to new software dev methodologies > Pre-2000s 2000s 2010s 2020s > Tech breakthrough PCs Mainframes, Web, client-server APls, mobile Cloud, Al coding assistants > methodologies Softwaredev Waterfall Agile dev Product and platform dev Al-nativedev > 3 > McKinsoy&Company ![[assets/reconstructed-slides/SZStlIhyTCY/slide-004.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-004.html) - AI slide classifier: `content_slide` confidence `0.98` - Text source: advanced OCR `rapidocr-live/bright-screen/contrast`. - OCR decision: ready — Dense slide plus lower-third title card and sponsor footer; OCR is better than manual transcription here. Slide text: > New technologies have given rise to new software dev methodologies > Pre-2000s 2000s 2010s 2020s > Tech breakthrough PCs Mainframes, Web, client-server Cloud, APls, mobile Al coding assistants > methodologies Software dev Waterfall Agile dev platform dev Product and Al-native dev > McKsay&Comgany > AIE/LEAD MOVING AWAY FROM AGILE: WHAT'S NEXT? > Google DeepMind PRESENTEDBY Senior Partner MARTINHARRYSSON Business Analyst NATASHA MANIAR McKinsey & Company ![[assets/reconstructed-slides/SZStlIhyTCY/slide-005.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-005.html) - AI slide classifier: `content_slide` confidence `0.99` - Text source: agent_vision. Slide text: > MOVING AWAY FROM AGILE: WHAT'S NEXT? > How can you go from "10x engineers" to a "10x team"? > How can you scale a "10x team" to a "10x company"? ![[assets/reconstructed-slides/SZStlIhyTCY/slide-006.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-006.html) - AI slide classifier: `content_slide` confidence `0.99` - Text source: advanced OCR `rapidocr-live/bright-screen/contrast`. - OCR decision: ready — Text-dense chart slide with axis labels, numbered bottleneck list, and footer logos. Slide text: > However, bottlenecks in processes may prevent high individual developer productivity to translate to equally high team productivity > Output(featuresper year) (due to increasing coding demand) model intelligence and increasing feature Output potential Bottlenecks amongteammembers Collaboration overhead > processing change Cognitive limits on > Bottlenecks (features delivered Output reality withintimeconstraint) Communication gaps among team members Manual review and debugging time > Increased complexity of > Years morecodegenerated > McKimeyCompa > AIE/ LEAD ex McKinsey Martin Harrysson &Company Katelyn Lesse ANTHROPIC EVERY Dan Shipper Asaf Bord Northwestern Mutual' Michele Catasta replit > Google DeepMind PRESENTED BY elastic Red Hat greptile comet CopilotKit Go gle Deep ![[assets/reconstructed-slides/SZStlIhyTCY/slide-007.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-007.html) - AI slide classifier: `content_slide` confidence `0.98` - Text source: advanced OCR `rapidocr-live/bright-screen/opencv-adaptive`. - OCR decision: ready — Diagram slide with multiple small labels and boxed annotations; OCR is likely better than manual transcription. Slide text: > Bottlenecks within current operating model and team setup > Botleneck > Dolaysfromincreasedcomplexoty and security vulnerabaites > Refinement. OayO DoyB Retro. > thard.to-interprot pue siodoponop! Inoiciont task. agonts duo to spocifications Ussignment among Sprint planning Doyl Development: Oay23 Sprint reviewz Day4 > Comeury > AIEngineer Engineering thefuture of Al ![[assets/reconstructed-slides/SZStlIhyTCY/slide-009.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-009.html) - AI slide classifier: `content_slide` confidence `0.99` - Text source: advanced OCR `rapidocr-live/bright-screen/contrast`. - OCR decision: ready — Dense multi-column operating-model slide with many small text blocks and labels. Slide text: > For each tech function, there may be a different operating model > Types of work Future example operating models > Modernization Humans supervise factory of agents modernising legacy continuously Agentic factory > Maintenance ticketswith minimal humansupervision Agents process lowest complexity > Brownfield products Greenfield products needs,generatedesigns,codeand tests Factories of agents discover customer with human supervision Al co-creator innovation lab > Infrastructure &operations Agents process lowest complexity tickets with high level of human supervision due to higher risk of impact on critical services Human-led with co.pilots > AlEngineer Engineering the future of Al ![[assets/reconstructed-slides/SZStlIhyTCY/slide-010.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-010.html) - AI slide classifier: `content_slide` confidence `0.99` - Text source: advanced OCR `rapidocr-live/bright-screen/contrast`. - OCR decision: ready — Text-heavy chart slide with small labels, multipliers, and outcome callouts suitable for OCR. Slide text: > Global survey of 300+ enterprises showing what differentiates top performers > performers X-Likelihood of shift compared tobottom 7x Shifts 6x Outcomes > (E2Eimplementaton ofAlof4+usecases) workflows Al native + (broader skillsets, Al native newroles) roles 5-6x Faster time to market 2-3weclo 1.2 weeks > Enablers > 2x 2x 7x 3-4x Higher quality artifacts > Upskilling measurement Impact Performance reviews > McAesry &Company > AIE/ LEAD MOVING AWAY FROM AGILE: WHAT'S NEXT? > Google DeepMind PRESENTEDBY Senior Partner MARTIN HARRYSSON Business Analyst NATASHA MANIAR McKinsey & Company ![[assets/reconstructed-slides/SZStlIhyTCY/slide-011.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-011.html) - AI slide classifier: `content_slide` confidence `0.99` - Text source: advanced OCR `rapidocr-live/bright-screen/contrast`. - OCR decision: ready — Dense comparison slide with multiple small cards and labels; OCR is the best triage path. Slide text: > Besides adopting tools what needs to shift? > Al native workflows Operating model: Al-native roles Talent: > Team ceremonies (continuousplanning with instead of two.pizza pod) Team size (one-pizza > Unit of work Team configuration > (spec-drivendevelopment instead of story-driven) from specialized roles) (definers and builders > Ways of working (code-based prototypes from long PRDs) (agent managers instead of specialized practitioners) Roles and skills > Monry & Compeny 10 > AIE/LEAD MOVING AWAY FROM AGILE: WHAT'S NEXT? > Google DeepMind PRESENTED BY Senior Partner MARTINHARRYSSON Business Analyst NATASHA MANIAR McKinsey & Company ![[assets/reconstructed-slides/SZStlIhyTCY/slide-013.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-013.html) - AI slide classifier: `content_slide` confidence `0.98` - Text source: advanced OCR `rapidocr-live/bright-screen/contrast`. - OCR decision: ready — Dense metrics, multiple labels, and small footer text are better handled by OCR. Slide text: > Operating model: Positive impact of interventions > Adoption Agent consumption, # average ACUs Active users of Al coding assistant tool > 60x vs end of Q2 +28% Highly active users vs. end Q2 > Speed Number of code merges, biweekly average per frontrunner squad > +51% Vs end of Q2 > Efficiency Business impact efficiency, avg hours per unit of work for each frontrunner squad > -60% vs. end o! Q2 -34% vs. avg Q3 > McKescy&Company 12 > AIE/LEAD MOVING AWAY FROM AGILE: WHAT'S NEXT? > Google DeepMind PRESENTEDBY SeniorPartner MARTINHARRYSSON Business Analyst NATASHA MANIAR McKinsey & Company ![[assets/reconstructed-slides/SZStlIhyTCY/slide-014.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-014.html) - AI slide classifier: `content_slide` confidence `0.99` - Text source: agent_vision. Slide text: > Talent: Top performers are creating new AI-native roles and pivoting responsibilities of existing roles > Top 3 roles most shifted by AI1, % of respondents > Software engineer 71% > Product manager 69% > Testing / QA 57% ![[assets/reconstructed-slides/SZStlIhyTCY/slide-015.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-015.html) - AI slide classifier: `content_slide` confidence `0.98` - Text source: advanced OCR `rapidocr-live/bright-screen/contrast`. - OCR decision: ready — Diagram with many small labels and role names; OCR will capture the team structure more reliably than manual transcription. Slide text: > and higher feature throughput in large international airline Talent: Smaller but higher number of teams enabling higher capacity > From > Pod # 1 PM Pod # 2 PM > Tech Lead 2 Toch lead 二 > 222 Devs Devs > QA 22 QA > Two pizza team of 8-10 people > Examplesharedfunctionalroles > Devops SREI TPM anslyst Business analysts/ Data engineers > MckGesay& > AlEngineer Engineering the future of Al ![[assets/reconstructed-slides/SZStlIhyTCY/slide-016.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-016.html) - AI slide classifier: `content_slide` confidence `0.98` - Text source: none. - OCR decision: ready — Two-column org-structure diagram with many small labels and role names is OCR-suitable. ![[assets/reconstructed-slides/SZStlIhyTCY/slide-017.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-017.html) - AI slide classifier: `content_slide` confidence `0.99` - Text source: agent_vision. Slide text: > Deeper organizational shifts required for large enterprises > Rewiring the champions > Change management > Rewiring 100+ teams ![[assets/reconstructed-slides/SZStlIhyTCY/slide-018.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-018.html) - AI slide classifier: `content_slide` confidence `0.98` - Text source: advanced OCR `rapidocr-live/bright-screen/contrast`. - OCR decision: ready — Chart labels, annotations, and the right-side unlock list are dense enough that OCR is the better first pass. Slide text: > Case study of scaling productivity in top 10 technology company 01010 10101 10101 > Top 10 technology company for long term behavior change Unlocks needed to scale adoption > 一GenAl tool DailyActive Usors(#)一DAU retention(%) > 450 400 350 000 250 200 during training Usage spiked weeks... Nr 35 30 25 20 40 45 and coaching Al and agents Hands-on upskilling Resetexpectations of > 150 100 50...but many unlocks sustain behavior were nceded to change 15 10 Spark grassroots movement > Oct Nov Dec Jan Feb Mar Ap May Jun Jul system Build measurement > McXrsoy &Company 18 > AIE/LEAD MOVING AWAY FROM AGILE: WHAT'S NEXT? > Google DeepMind PRESENTED BY Senior Partner MARTIN HARRYSSON Business Analyst NATASHA MANIAR McKinsey & Company ![[assets/reconstructed-slides/SZStlIhyTCY/slide-019.jpg]] - Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/reconstructed/SZStlIhyTCY/slide-019.html) - AI slide classifier: `content_slide` confidence `0.95` - Text source: advanced OCR `rapidocr-live/bright-screen/contrast`. - OCR decision: ready — Dense mixed-layout slide with diagram labels, small body text, and a stage-crop composition that is better handled by OCR than direct transcription. Slide text: > Holistic measurement system to evaluate enterprise-wide change > Outcomes > Velocity Capacity Security Quality Resiliency > sndnos Breadth and depth of adoption Number of people upskilled Developer NPS and attrition rate > Inputs SinvestmentinAcoding/devtools Sandtimeinvestedintraining/ upskilling programs Sand time inchange managemeny op model ransformation > McKrsny & Company 21 > AIE/ LEAD MOVING AWAY FROM AGILE: WHAT'S NEXT? > Google DeepMind PRESENTEDBY Senior Partner MARTIN HARRYSSON Business Analyst NATASHAMANIAR McKinsey & Company ### Hidden Non-Slide Evidence - [`slide-001.jpg`](/assets/reconstructed-slides/SZStlIhyTCY/slide-001.jpg) — `speaker_stage` confidence `0.9`; Stage shot with presenters and conference signage, not a readable presentation slide. - [`slide-008.jpg`](/assets/reconstructed-slides/SZStlIhyTCY/slide-008.jpg) — `speaker_stage` confidence `0.99`; Stage camera shot with speakers on stage, not a presentation slide. - [`slide-012.jpg`](/assets/reconstructed-slides/SZStlIhyTCY/slide-012.jpg) — `other` confidence `0.99`; Audience shot, not a presentation slide. - [`slide-020.jpg`](/assets/reconstructed-slides/SZStlIhyTCY/slide-020.jpg) — `title_card` confidence `0.99`; brand/title splash, not a substantive presentation content slide Classification audit: `raw/sources/slide-ai-classification/reconstructed/SZStlIhyTCY/audit.json`