Try searching for

Portfolio Presentation

1 / 136
Read the slide text
  1. Furquan Ahmad — Portfolio Presentation · Portfolio presentation
  2. My Path To Today
  3. My Path To Today
  4. My Path To Today
  5. My Path To Today
  6. My Path To Today
  7. My Path To Today
  8. My Path To Today
  9. Lead Product Designer at Scale AI — Selected work: Shour, automating cases for lawyers at Qatar's Ministry of Justice · Before: Senior Product Designer, Meta
  10. UX Research Photos — Content, community and travel board
  11. My Content — Journey · Publish a design newsletter and blog with · every month. · Featured Figma Community contributor · with · Design YouTube channel with · Mentored over · designers into jobs. · 1000+ readers · 50+ · 100,000+ · views. · 100,000+ downloads. · Making a difference · Furquan
  12. Community leader and culture builder — Building Culture · Crescent Creatives - · Founder of 100+ · member Muslim · design community. · Muslims @ Meta · London Lead running · 15+ events with 1000s · of attendees. · Led initiatives raising · Workplace design · team's pulse score · 42% (2023-2024). · Design culture · catalyst. Led the · running of design · rituals across · Workplace.
  13. Shour: automating cases for lawyers — AI drafts cited defence memos from case documents and speeds up the legal process · Designed and built the entire frontend
  14. Shour live prototype
  15. UX Research Photos — Projects at Meta · Projects at Meta
  16. Case Study — Rethinking · Workplace To Be AI · First. · The Dawn of Workplace AI: Launching Meta's · First Intelligent Feature.
  17. Overview — 01 · 02 · Problem & · Process · 03 · Solution · 04 · Execution · & impact
  18. What is Workplace? — Workplace is like a private social · network for your company. · It feels like regular Facebook, but it's · . · It combines familiar social features with · professional tools, enabling employees · to chat, share files, and collaborate. · It's great for building a sense of · and keeping · everyone · . · designed for work · community at work · connected · Workplace explained
  19. Gen AI Landscape — + Opportunity · New wave of Gen AI tools reshaping · industry standards. · AI features now · considered table · stakes. · The overall AI in workplace market is · projected to grow from · Generative AI could add the equivalent · of $2.6 trillion to $4.4 trillion annually to · the global economy by the year 2040. · $4.8 billion in · 2022 to $33.5 billion · by 2030. · Background · Source:
  20. from vision to execution — Led design · Defined · & · presented to · leadership. · strategy · vision, · 3 · year · Mentored · on · execution for Gen AI · work. · 4 other · designers · Established Meta’s · Gen AI · for work. · design · principles · first · Collaborated across 4 · orgs with different · XFN to establish · principles. · My role
  21. 01 — Overview · Problem & · Process · 02 · 03 · Solution · 04 · Execution · & impact
  22. Event Organisers — Workplace AI · JTBD · is · when searching for content · more efficiently · and confidently. · content efficiently. · Harnessing data · Making sure information · , and knowing · has · understood the information. · easy to · understand · who · Saving time · Creating content · Moderating · insights. · Pixie Zhang - Event Organiser
  23. Knowledge Workers — Workplace AI JTBD · I want to feel informed and up to date, I · want to · catch up on key · information from company broadcasts, · but I · to watch all · of them. · This is a need to control workload, keep · on top of ongoing projects and · on relevant official information. · efficiently · don't have enough time · stay up to · date · Waleed Ibara - Viewer
  24. BUSINESS PROBLEM — How might we combine Workplace’s strengths in · org comms and community with an innovative AI · vision to retain high-potential Workplace · customers? · Business problem
  25. Validate through — interviews · Putting projects on the roadmap · 90-minute remote semi-structured interviews. · Participants were asked about their · general perceptions of GenAI, previous · use of GenAI in their personal lives and for · work, pain points & frustrations with · Workplace, and feedback & perceptions · of new concepts. · Concept feedback and ranking showed · high agreement across all user segments.
  26. Content creation and search — Personalisation and quality · Data privacy · UXR AI Customer Problems · Organisers & Hosts · Pixie Zhang - Event organiser
  27. UXR + PMM Customer Problems — Viewers · Discoverability problems · Hard to feel a connection · Difficult to catch up · Waleed Ibara - Viewer
  28. Leading workshops — Putting projects on the roadmap · Focused on Gen AI integration for Workplace: · Defining current AI features and · capabilities. · Assessing user needs and · expectations. · Led key spicy debates on long-term AI · strategy: · Balancing AI presence with content · quality. · Determining appropriate moderator · control.
  29. Vision to shipped — Putting projects on the roadmap · Led the creation of a · A strong focus on harnessing Gen AI to · solve customer problems. · This vision · on · different teams’ roadmaps that brought · about transformative change. · 2-year design vision · to guide platform direction · sparked multiple projects
  30. LIVE 2D Engagement Vision
  31. 01 — Overview · 02 · Problem & · Process · Solution · 03 · 04 · Execution · & impact
  32. From Vision to Hack — of AI · Chapters solves customer · problem: Enables easy navigation · through long videos. · Supports Team Goals: Enhances video · engagement and accessibility. · Goal: Achieve · by · demonstrating practical benefits · during the hackathon. · concept buy-in · Hackathon
  33. Hackathon project demo
  34. 01 — Overview · 02 · Problem & · Process · Solution · 03 · Execution · & impact · 04
  35. How we chose automatic chapters for execution? — Lower the barrier for · customers to seek · valuable knowledge · Retain existing · Workplace customers · who actively use · Events. · Improve the top line · metric centred · around average · watch time. · Events is a big rock · investment for · Workplace providing · direct value to SBRs. · Lower the barrier for · customers to seek · valuable knowledge. · Retain existing · Workplace customers · who actively use · Events. · Improve the top line · metric centred · around average · watch time. · Events is a big rock · investment for · Workplace providing · direct value to SBRs. · Mapping to OKRs
  36. Product Requirements For Chapters — Viewer flow · Priority · View  chapters whilst scrubbing on · the video scrubber, on newsfeed · and fullscreen. · P0 · View a summary of what was · announced in each individual · chapter from the Workplace post. · P0 · Consume individual chapter · summary whilst scrubbing on the · video player. · P1 · Consume thumbnail  of video · chapter from video scrubber. · P1 · Moderator flow · Priority · Entrypoint in 3-dot menu to access · chapter generation from Workplace · post. · P0 · Enable auto-generation of chapters · CTA. · P0 · Once chapters have been · generated you can get a further · summary of what was announced · inside individual chapters. · P0 · Consume thumbnail  of video · chapter from video scrubber. · P1 · Mobile flow · Priority · View chapters whilst scrubbing on · the video scrubber on fullscreen. · P0 · View a summary of what was · announced in each individual · chapter from the Workplace post. · P0 · Consume individual chapter · summary whilst scrubbing on the · video player. · P0 · Consume thumbnail  of video · chapter from video scrubber. · P1 · Mapping out the feature list
  37. Measurement — Framework · % of meaningful videos that are with · video chapters (Adoption). · . · . · . · % Auto chapters users that have used the · auto-chapters 2+ times (User value) · Auto chapters enable/disable ratio (User · value) · % videos where the user had to correct · the auto-generated chapter text · Tracking impact
  38. Design Deep Dive — Automatic Chapters for · Viewers and Organisers · Diving into the designs
  39. Newsfeed — What do the interactions look like · with the summary open? · How does the chapter look when · you’re playing the video on · newsfeed? · Waleed Ibara - Viewer · Exploration
  40. Chosen Solution — Summary always visible for more efficient consumption after internal dogfooding feedback. · Combining the · chapter time and · pause inside a · single button. · Summary is always · capped to 2 · sentences for · accessibility
  41. Waleed Ibara - Viewer — Walkthrough
  42. Upselling — How can we make organisers aware · chapters have been generated for · their video? · How can we enable event organisers · to efficiently add chapters to their · videos? · Chapters always have to be reviewed · before others can see them. · Pixie Zhang - Event organiser · Walkthrough
  43. User education — Design plays an integral part in · ensuring that people have a positive · user experience across MetaWorks · when using an AI product. · is key. · For many people, Workplace could be · the first place they encounter GenAI · and · how to use it for · their day-to-day work · understanding · Pixie Zhang - Event organiser · Walkthrough
  44. Pixie Zhang - Event organiser — Full control · Walkthrough · You have the ability to edit any title · or summary generated or change the · start and end time. · We know models can hallucinate so · the chapters produced won’t always · be perfect.
  45. The GTM plan — Shipped to company wide dogfooding. · Shipped to a public alpha release. · Ship to GA. · Working with PMM on GTM plan
  46. After internal dogfooding feedback we — shipped the product to external customers · in an open Alpha release. · Customers were excited to try the new · Gen AI features and understood the value it · could bring in their current workflow. · Alpha feedback · Shipping to public
  47. Impact and Results — Proportion of viewers interacting with the · AI-generated chapters is · than for manual chapters, likely due to the · videos with AI chapters · This indicates that the users might receive · more value (easier for the creators to · create more chapters, easier for the · viewers to navigate the videos in their · consumption sessions). · 🚩Quality improvements needed, 15-17% · users experienced critical errors in chapter · creation (target <10%). · slightly higher · having more · chapters on average. · Metric · Value · Target · % video creators used video · chapters. · ✅ ~13%. · >3% · % videos with chapters that are · auto generated. · ✅ ~82% · >50% · % Auto chapters users that have · used the auto-chapters 2+ times. · ✅ 30-40% · >19% · % of viewers interacting with · chapter (selecting, sharing) vs · total viewers of videos with · chapters. · ✅ 4-6% · >5% · Creation side: % videos where the · user had to correct the auto- · generated chapter text before · publishing. · ✅ 18-22% · <50% · % users with critical error in · chapter creation. · ❌ 15-17% · <10% · Metrics
  48. Gen AI For Work — PRINCIPLES · Gen AI Principles For · Work
  49. Creating the — principles · Partnered with teams · who · are involved in Gen AI projects to develop · these principles. · Used our learnings from our workshops · and shipping the first Gen AI feature · Chapters for work. · When designing for work, there’s · company-wide · increased accountability for data and · sensitive information. · Gen AI Principles
  50. Sharing learnings — across Meta · Shared the learnings by creating a · Workplace blog series which received · Spoke internally at different conferences · about how to use the principles. · execution · These principles have informed the · for · 4 different projects across · Workplace. · 10,000+ reads internally. · Shaping roadmaps
  51. Informing different — Roadmaps · audio, · . · These principles have · for · , · including chapter summaries, composer · assistant, · post and comment · summarisation. · on utilising Gen AI · on their projects · Presented these principles during SBR · reviews to support Workplace contract · renewals. · informed the · roadmap · 5 different projects · Mentored 4 designers · Shaping Workplace roadmaps
  52. Slide 49
  53. Slide 50
  54. Slide 51
  55. Core Principles — Viewers · Lead with the problem and · add value · Anticipate risk and lead with · human reviews · Don’t oversell AI and · compromise over features · Waleed Ibara - Viewer
  56. Learnings and — Next Steps · Give creators · over the output · of generated chapters to help build · confidence. · such as · the video transcript. · Provide · /transparency · from the AI output to build trust · Work more closely with the legal teams to · streamline reviews and understanding of · AI. · more references · more control · Learnings
  57. Case Study — Growing Facebook · Groups · Redesigning discovery and recommendations for · Groups. Experiments delivered consistent 2–7% lifts, · adding hundreds of millions of joins across surfaces.
  58. Overview — 01 · 02 · Problem & · Process · 03 · Solution · 04 · Execution · & impact · Overview
  59. Facebook Groups — Facebook’s mission is to give people the · power to build · and bring the · world closer together. · Groups is central to this mission, · enabling people to connect over · Facebook has over 3 · billion monthly active users, with · Groups playing a central role in how · people interact on the platform. · More than 1.9 billion · people use Groups every day, making it · one of the most widely used features · across Facebook. · shared · interests. · Global scale: · Daily engagement: · community · Background
  60. My team — My team focuses on improving FB Group Discovery and join conversions across the Meta family of apps · Feed, Marketplace, Reels · Community · Feed, Stories · Background
  61. to grow Facebook Groups — 6 month project · Established new · recommendation patterns · Landed recommendations · on 6 new surfaces · Unified all units to increase · optimal join conversions · My role · 20+ · experiments · 6 · new surfaces · +5% · join conversions
  62. from vision to execution — Design lead · Defined · & · with EM and · PM presented to · leadership. · strategy · vision · Led complex · collaboration across · 3 orgs - Facebook, · Instagram & · WhatsApp. · Sole designer on · Groups · Recommendations · team. · Shipped changes for · FB Groups · Recommendations · that increased join · conversions by · 5%. · My role
  63. 01 — Overview · Problem & · Process · 02 · 03 · Solution · 04 · Execution · & impact · Problem and Process
  64. JTBD Young Adults — Facebook JTBD · When entering · the workforce, moving cities, or setting up · my first home, I need a single platform · that helps me manage practical tasks. · Only 20% of · YAs add a new friend · on · Facebook. Instead, Groups and · Marketplace drive the majority of · engagement. · With friend · growth plateauing, Groups provide a · community feel. · Navigating milestones: · Shifts in social behaviour: · Community over friending: · after 30 days
  65. PROBLEM STATEMENT — How might we help young adults find · based on their personal interests · when using Meta products in an intuitive way? · relevant · communities · Problem
  66. 3 Problem Areas in Group Recommendations — User research shows young adults (YAs) struggle with group recommendations because they: · They don’t understand why groups are suggested (missing context and signals). · Key entry points for YAs aren’t surfaced in the right places. · Layouts and styles vary across units, lowering join conversion. · Lack Clarity – · Missed Reach – · Inconsistent Design – · Background · Lack of Information · Missed Reach · Inconsistency
  67. Recommendations are not informative or optimised — 20+ experiments in 6 · months · 1 product designer, 12+ Eng, 1 · PM, 1 Data Scientist, 1 EM · Increased join conversions across · recommendations by 5%. · Released: · 1 team:
  68. Problem 1: Not meeting YAs at surfaces they use — Background · Yusuf - Young adult · Groups appear in Feed, Marketplace, Stories, Groups Tab · and Reels, but placements often miss the moments that · matter most for YAs. · Entry points are missed, leading to lower engagement at · key moments. · Many YAs use Instagram without Facebook, creating major · reach gaps. · Existing placements feel peripheral (e.g. buried in · Marketplace or mid-card in Reels). · Optimising Marketplace and Reels visibility would bring · Groups closer to core experiences. · Expanding into Instagram Feed and Stories would extend · reach beyond Facebook. · Meeting young adults where they already spend time makes · Groups feel more relevant. · f · Facebook · Where Groups already appear · Feed · Stories · Groups Tab · Marketplace · Reels · ← optimise visibility · Instagram · Where many YAs are, often without Facebook · Feed · Stories · ← new entry points
  69. Problem 2: Inconsistent join experience — Background · Inconsistent patterns in the join units: varying · content, different image sizes, button positions, · metadata and social proof labels. · This meant some units were naturally getting better join · conversions than others. · We hyper-optimised the recommendation cards throughout · Facebook. · Years of data from past recommendation units showed us · what performs well and what needs work. · We ran 20+ experiments to find the best content, button · positions, badges, labels and image heights, and when to · change the header type. · Join · Badges · Image height · Header type · Metadata · Social proof label · Button position · What we tested across the units · 20+ · experiments in 6 months
  70. Problem 3: Lacking context and social proofing — While friend-based social proof remains a strong driver of group joins, it’s · often unavailable (8% of times) especially for niche or new groups. · To maintain trust and relevance in these moments, we’re expanding the · scope of social context beyond mutual friends. · This includes surfacing interest-based tags (e.g. For solo travellers, · Popular for hiking), geo-relevant cues (e.g. Popular in San Francisco) and · discussion topics inside the group. All of these help signal personal · relevance and build user confidence at a glance. · These updates align with our broader standardisation efforts across all · GYSJ modules: simplifying headers (+2.71%), fixing padding (+4.7%), · and surfacing social context (+3.94%) all contributed to measurable lifts · in joins. · Background · Today: friend-based social proof · Expanding to interest and place cues · For solo travellers · Popular for hiking · Popular in San Francisco
  71. AI Opportunity For Groups — AI can power smarter recommendations by summarising group discussions, explaining why a group is relevant to · you, and building confidence in why you’d want to join. · Background · Summarise Group Content · Explain Group Relevance · Predict & Boost Engagement
  72. 01 — Overview · 02 · Problem & · Process · Solution · 03 · 04 · Execution · & impact
  73. Upselling on FB Reels — Reels is one of the most popular surfaces for YAs, with · short-form content driving high engagement and · repeat visits. · By using the Reels graph, we can match people’s · interests with relevant Group recommendations. · Signals such as topics watched, location, and mutual · connections help us personalise results and make · them feel more relevant. · Example: Someone who frequently watches running · videos may see local running groups suggested. · The aim is to improve discoverability, relevance, and · engagement. · Problem 1: Not meeting YAs · Yusuf - FB Young adult
  74. IG Stories and Feed — A significant portion of YA IG users don’t · use Facebook. By partnering with the IG · team we can match their IG interest graph · to pair with the FB graph to produce Group · recommendations inside the feed and on · stories. · As this is on IG Feed the unit space to run · the experiment cost millions so we · partnered with the Marketplace team to · add strength to running the experiment. · Problem 1: Not meeting YAs · Yusuf - FB Young adult
  75. Improving consistency — Problem 2: Inconsistent join experience · Data showed that 95% of group names are under 15 · characters, so we made a principle of single-line · headers to cut empty space. · Larger image headers led to better conversions, so we · reduced the meta text space and increased the image · height across all units. · Proposal: add social tags to media cards · We proposed adding social tags (AI-generated keywords) · as information chips on the media card. They help · people understand what a group is about, which matters · because ~90% of impressions don’t include friend-based · social context. · Proven optimisations in the new card · Lift in joins from past experiments · Visible “Remove” button · +9.67% · Fixing padding and layout · +4.7% · Simplified headers (1 line) · +2.71% · Larger group title size · +2.2%
  76. Principles we made — 1. Design for passive scrolling · People scroll in a low-effort, low-intent state, so the · unit has to work at a glance. Keep layouts clean, · headers short and cards visually light. · → Simpler headers (+2.71%) and fixed padding (+4.7%) · both improved joins. · → Backed by cognitive ease. · 2. Prioritise personal relevance · People only pause for what feels personally meaningful. · Lead with bold group names, engaging images and clear · social context like “3 friends are members”. · → Social proof led to +3.94% more joins; 2-line group · names (+2.2%) helped with clarity. · → Backed by relevance heuristics and social proof bias. · 3. Guide to one clear action · Reduce choice overload with one consistent button: Join. · Play down secondary actions and keep the button in the · same spot to build habit. · → Playing down “View” led to +0.5% more joins; · structured footers (+0.57%) performed better. · → Backed by choice overload and decision simplicity. · 4. Respect control and exit intent · Let people dismiss irrelevant groups with a clear · “Remove” button. Control makes them more likely to trust · and engage with what remains. · → A visible Remove button led to +9.67% more joins · and cleaner impression quality. · → Backed by perceived control and autonomy. · Problem 2: Inconsistent join experience
  77. New card we designed — Problem 2: Inconsistent join experience · One card for every recommendation unit, built on what · our experiments showed works. · Single-line group name · Simpler, one-line headers lifted joins by 2.71%. · Larger image · Bigger image headers led to better conversions. · Social context · “Join Maryam and 20+ friends” shows who is already · there. Social proof led to 3.94% more joins. · Signs of an active group · Member count and posts a day, on one short line. · One clear action · A single Join button. Playing down “View” added 0.5%. · Visible Remove · An X to dismiss the group lifted joins by 9.67%.
  78. Problem 3: Lacking social proofing — Background · Key data point · In past Group You Should Join (GYSJ) experiments, social · proof (e.g. “3 of your friends are in this group”) led · to a +3.94% uplift in joins. · Why it works · Social validation lowers the cognitive load for people. · It builds trust quickly, especially when people are · scrolling passively and not searching for a group. · However, mutual friends are available in fewer than 10% · of impressions, which limits scale. The impact stays · strong when they are shown. · +3.94% · uplift in joins from social proof · <10% · of impressions have mutual friends to show · How it shows up on the card
  79. Using AI to Scale Contextual Signals — Background · AI can generate: · Keyword labels from group descriptions and post content · (e.g. “Travel Tips”, “South Asian Parenting”, “Finance · for Gen Z”) · Trending or popularity tags (e.g. “Popular in SF”, “3K · members joined this week”) · Local relevance signals (e.g. “Based in Oakland”) · Topic summaries from group posts using LLMs · Tone detection (e.g. humorous, serious, community · support) · Related experiment result: in Snapshot Preview, adding · AI-generated topic keywords increased post-join · engagement by 1.73%. · AI-generated labels · Travel Tips · Popular in SF · Based in Oakland · 3K members joined this week · AI summary of group posts · Join · +1.73% · post-join engagement with AI · topic keywords (Snapshot Preview)
  80. Iterations — Problem 1: Not meeting YAs · Both start from a link under a running Reel, then show groups that match it. · Option A · Full page of groups · A link under the Reel opens · a full page of related · groups to browse and join. · Option B · Groups in a sheet · The link opens a sheet over · the Reel, so people can scan · running groups and join · without losing their place.
  81. 01 — Overview · 02 · Problem & · Process · Solution · 03 · Execution · & impact · 04
  82. Execution & impact — Execution & impact · +5% · join conversions across · recommendations · 20+ · experiments released · in 6 months · 6 · new surfaces with Group · recommendations · +2% · click-through from AI · group summaries · What moved joins on the new card · Visible Remove button · +9.67% · Fixing padding and layout · +4.7% · Social proof · +3.94% · Single-line headers · +2.71% · Larger group titles · +2.2% · Where it landed · Group recommendations now meet young adults · in Reels, Feed, Search, Group Mall, WhatsApp · and Instagram, using one shared card pattern. · AI topic keywords in Snapshot Preview lifted · post-join engagement by 1.73%. · I presented the AI recommendation formats · to VP-level stakeholders.
  83. Case Study — Post-Production · Studio · Designing a brand new studio that turns Live · Broadcasts into compelling, shareable · content.
  84. Project Background — Project background
  85. Event Organiser — Post Event JTBD · Making sure information · , and knowing · has · understood the information. · Creating a record of the broadcast to · share highlights with those who missed it · or want a summary. · Analyse engagement metrics to identify · areas of high interest and improve future · broadcasts. · easy to · understand · who · Pixie Zhang - Event Organiser · Pixie Zhang - Event organiser
  86. There’s — to manage your · broadcast. · The current experience is · and · inconsistent, you have individual entry points · which all look different. · for the current experience (<3% · for certain tools) · , only the person that started · the broadcast has access to these tools. · Once you exit the broadcast you can’t reach all · the tools again. · fragmented · Adoption is low · Permission issues · no central place · Experience Today · Post production today · Pixie Zhang - Event organiser
  87. Post production problem areas — Organisers & Hosts · Experience misalignment · Content Chaos · Permission issues · Pixie Zhang - Event organiser
  88. Led design and product from strategy to execution — Defined strategy & · vision and presented · to leadership. · Lead designer for · post production · space creating a new · surface in 6 months. · Created a new stream · of work for future · Workplace projects. · Collaborated across · multiple orgs with · different XFN. · My role
  89. PEOPLE PROBLEM — HMW enable organisers to effectively manage · their live broadcasts after the event has finished? · People problem
  90. Defining Success — Success for the project
  91. Key principles — Unified Experience · Create a seamless experience across all post- · production tools. No matter which tool you're using, it · should · with a familiar design language. · feel like home, · Flexible Control · Give organisers the freedom to manage who can help · with their live video. They can · to join in · and assist, but only with their permission. · invite others · Streamlined Efficiency & Confidence · all tools into one platform to streamline · workflow and boost user confidence in managing and · editing videos effectively. · Centralise · Effortless creation & sharing · Enable organisers to swiftly manage and edit key · broadcast segments, facilitating · to · achieve optimal viewer engagement. · easy sharing · Product principles
  92. Feature — Milestone · Requirements · Post-production entrypoint · M1 · Making it as easy as possible to access post production tooling from · different Workplace entry points. · Design new studio foundations · and layout · M1 · Create a centralised studio hub where you can access all post production · tooling. · View Post · M1 · Give moderators the ability to see the Workplace post the video is · attached to. · Delete Video · M1 · Give moderators the ability to delete the broadcast. · Chapters · M1 · Create a new chapters experience that lives inside the new post · production studio. · Trimming · M1 · Improve the existing trimming experience. · Clipping · M2 · Improve and import over the clipping experience. · Captions · M2 · Improve and import over the caption experience. · Product Requirements
  93. Metrics for Success — Human Language Goal · Metric · Maintain or increase · the Post production · tools adoption rate · # and % of real live streams created with at least one post production tool used · via Live Studio for Streaming Software vs Live Producer for Streaming Software. · Click-through rate (clicks out of impressions) on at least one post-production · tool by a Host using Live Studio for Streaming Software vs Live Producer for · Streaming Software. · Maintain or increase · the chapters adoption · rate · % of real live streams with Chapters created via Live Studio for Streaming · Software vs Live Producer for Streaming Software. · # of Chapters creation initiated (click-through rate). · % Chapter creation completed (successfully saved). · Maintain or increase · the Trimming adoption · rate · % of real live streams with Trimming created via Live Studio for Streaming · Software vs Live Producer for Streaming Software. · # of Trimming creation initiated (click-through rate). · % trimming creation completed (successfully saved).
  94. Explorations For Studio — Studio explorations
  95. Tools audit — Pixie Zhang - Event organiser · Data and research insights · Conducted a comprehensive end-to- · end audit of existing tools. · Leveraged existing data and UXR · insights to pinpoint key gaps. · Example: Clipping feature usage · below 2% due to entry point and · permission issues. · Addressed issues to improve user · experience and increase feature · engagement.
  96. Crafting an intuitive and streamlined IA — Pixie’s goals after a live finishes: · Focus on post production needs · Making sure the right people can see the post-live broadcast · Gather insights to improve the broadcast · Live Engagement · Invite collaborators · View post · Analytics · Post production · Chapters · Trimming · Clipping · Captions · Delete video · Chat · Polls · Q&A · Stream metrics · Event logs · Activity Logs · Studio
  97. Optimising IA Layout for efficient post-production — Easy access to · other tooling on · the sidebar · Top navigation bar for · high level navigation of · Studio · Flexible column sizes to · focus on different tooling · needs · Pixie Zhang - Event organiser
  98. Streamlined IA Layout for Enhanced Tooling — Column sizes can · change depending on · the focus.
  99. Publish and save any edits — from the same place on all · tooling pages · Understand what tooling page you’re on. · Multiplayer collaboration · with others · Easy navigation back to the studio homepage. · Simplifying Navigation with Streamlined Toolbars
  100. Pixie Zhang - Event organiser — IA Explorations · Displaying all the menu items in the LHC. · This added a lot of noise and reduced efficiency · and focus on post production needs. · Trying to fit the post production tooling inside the · same container is limiting · Walkthrough
  101. IA Explorations — Putting post production tooling under one central · tab. · Each navigation item takes you to a dedicated · tooling page inside the Studio. · Having post production tooling open by default. · Pixie Zhang - Event organiser · Walkthrough
  102. IA Explorations — Hiding all the live tooling under one tab to make it · easier to focus on post production tooling. · Making the tooling items more simple and easier · to read and understand its purpose. · Pixie Zhang - Event organiser · Walkthrough
  103. Progressive Disclosure: Scalable — sidebar with post-production · focus. · List Items: Scalable and adaptable · to different statuses, with added · text and imagery for clear tool · understanding · Multiplayer, see who you can collaborate with, · opening up permissions. · See any of the · chapters added to the · video. · Quickly navigate to the · post associated with the · video. · Quick insight into any · analytics to help you · improve your broadcast. · Chosen Solution
  104. When tabbing having the editor — tools at the forefront makes this a · lot more streamlined · Text is expandable for · longer languages · Using a mixture of · images and texts · Using the design system · colour palette to make · sure the contrast is AA · compliant · Accessibility and localisation
  105. End to end flow walkthrough
  106. Design System — Expanding the · design system
  107. Making questions and comments — easier to differentiate with the · ability to upload questions · Making post production easier by · marking comments as official · answers · New play button with · timestamps for efficient · consumption · New status pills inside list · items to indicate · New scalable components · Chapters · Highlight key moments from your broadcast. · Published · New video player with chapter · segmentation and chapter titles for · efficient consumption
  108. Explorations — Striving For · Quality
  109. Creating consistency across all tooling pages — Production · Design
  110. Championed design fidelity without — QA: · Conducted regular design audits in · test environment. · Ensured production closely · matched Figma designs. · Prioritised quality bugs with XFN. · Created detailed engineering tasks · for each discrepancy. · Maintained Figma documentation · for all identified issues. · Developed comprehensive Google · Sheet to track all bugs. · Quality improvements · Quality first
  111. Live Studio in production
  112. Studio — Tooling · Explorations
  113. Adding chapters — Walkthrough · Pixie Zhang - Event organiser · Effortless Video Replay: Seamlessly replay videos · while adding chapters for enhanced navigation. · Informative Empty State: Educate organisers on · the benefits of chapters, highlighting how they · facilitate easier video navigation for viewers.
  114. Showing Chapters — Walkthrough · Pixie Zhang - Event organiser · Making it clear to viewers what chapters have · been added with the relevant title and timestamp · before saving the changes for others to see.
  115. Clipping a video — Walkthrough · Pixie Zhang - Event organiser · Consistent Design for Video Clipping: Utilise the · left panel to create clips, maintaining the same · design language as chapters. · Easily review added clips in the top right before · saving.
  116. End to end flow walkthrough — Pixie Zhang - Event organiser
  117. Impact and Results — Impact · ✅ Launched to company wide dogfooding. · ✅ 99% of SBR customers renewed after using the new tooling and seeing · the vision. · ✅ Launched to closed beta release. · ✅ Post production tooling adoption has increased. · ✅ Increase in adoption across all post production tooling thanks to more flexible · permissions, better entry points and improved user experience
  118. Learnings and — Next Steps · Data- · Decision Making: Don't dismiss tools · with low usage; they may still hold value. · Universal Post-Production Solution: Scale post- · production capabilities to accommodate all videos · across Workplace. · Automate post-production tooling with more AI · efforts. · Prioritise Quality: Integrate quality measures early in · the execution process · Collaborative Design: Engage the full XFN team to · champion your vision and drive collective ownership. · Informed · Learnings
  119. Feedback about me from peers — Furquan has been one of the · - driven non- · analytics XFNs I've interacted with in my time in Workplace. · most data- and research · Christina Tatli - Data Scientist · Furquan is one of the · most inclusive team members I've ever had in my · career. · Jessica Lee - Product Manager · Furquan's productivity this half has been impressive, as he has worked · on many projects, in parallel. As a result, he was able to unblock the · team and be one of the · main factors for the success of the overall · team. · Saar Bendoiz - Engineering Manager · Furquan demonstrated · . He has · deeply engaged in the discussion for duplicate category entry point · change. · how well he embraced ambiguity · Arwa Arif - Engineer · Furquan concisely communicates progress and action he needs · from XFN and Eng. As a result, eng was never blocked by design as · Furquan · and his timelines are communicated. · shows accountability · Fang Matchima - Engineer · Furquan is · designers I’ve worked with, he proactively · identifies important problems. He demonstrates great insight and · reliable judgment, and proactively strengthens partnerships. · one of the best · Ty Gallop - Content Designer · XFN feedback · Peer feedback
  120. Thank You! — Thank you!
  121. Appendix — Appendix
  122. Appendix: Trimming a video
  123. Appendix: Deleting a video
  124. Appendix: Newsfeed entry point
  125. Appendix: Broadcast ended
  126. Newsfeed — How can we incorporate chapter · markers and content summaries within · the existing newsfeed post container? · What's the best way to implement this · using the Blue design system? · How can we clearly indicate the current · chapter to viewers? · What visual cues or design elements · will help users track their progress · through the video? · Waleed Ibara - Viewer · Exploration · Exploration
  127. Tahoe view — How can we help viewers quickly · catch up on missed content while · watching videos in fullscreen? · How can we make it easy for viewers · to follow along with chapters and · access summaries in real-time? · Waleed Ibara - Viewer · Waleed Ibara - Viewer · Exploration
  128. Post Production — ing all your chapters from the post · production tooling, embedded with the · existing experience. · Cancel the generation at any point to give · users full control and trust. · Manag · Pixie Zhang - Event organiser · Walkthrough
  129. Pixie Zhang - Event organiser — Notifications · Walkthrough · Notifying a moderator to let them · know chapters have been generated · for their videos.
  130. Pixie Zhang - Event organiser — Generated Chapters · Walkthrough · Clearly displaying the chapter title · and summary. · Any chapter can be edited or deleted · before being published to give · moderators full control.
  131. Regenerate — Pixie Zhang - Event organiser · Walkthrough · You can try and regenerate chapters · and get a different response if you · weren’t happy with the first output. · We know models can hallucinate so · the chapters produced won’t always · be perfect.
  132. 1. Lead with the — problem · Super Speed: · Helps the user perform tasks t · how to do, but faster. For example, · automatically adding chapters to a video. · Inspire & Unblock: · Helps the user get started (combating · blank post/page syndrome with a rough · draft) or to explore possible directions. · Help Me: · Use Gen AI to guide them or show them · things they don’t know how to do. · hey already · know · Super Speed: · Helps the user perform tasks t · how to do, but faster. For example, · automatically adding chapters to a video. · Inspire & Unblock: · Helps the user get started (combating · blank post/page syndrome with a rough · draft) or to explore possible directions. · Help Me: · Use Gen AI to guide them or show them · things they don’t know how to do. · hey already · know · Principles
  133. 2. Create Familiarity — A Gen AI experience should · across product tools and clearly · communicate that AI is being used to the · user. · By following the same patterns and · guidelines, we can help achieve familiarity · and · and build · confidence in using our Gen AI products · for work. · look cohesive · reduce the learning curve · Principles
  134. 3. Anticipate risk — and lead with · human reviews · Inaccuracies and · can · have major implications, particularly in · a work context. · Give creators a · – whether this means providing · the option to review, redo, or delete · generated content. · Any generated content published to an · audience should be · . · human reviewed · hallucinations · significant level of · control · Principles
  135. 4. Don’t oversell AI — Gen AI is a new feature that everyone · wants to try and implement in their · product. · It should be part of a unified · experience that · other · users use. · doesn't compromise · existing features · Principles
  136. 5. Keep users — informed and build · trust · Generative AI is a new technology not all · users know a lot about, particularly how it · works and what data is being used. · early on · in user flows and let users know exactly · how the AI works, what information is · being used, and how it could help make · their day-to-day tasks easier. · Build a · and · in our users, especially for work use cases · where creators need to be mindful of the · information · Identify moments for education · high level of trust · confidence · Principles