{"id":8106810,"date":"2026-04-07T08:35:56","date_gmt":"2026-04-07T15:35:56","guid":{"rendered":"https:\/\/agilealliance.org\/?post_type=aa_event_session&p=8106810"},"modified":"2026-04-07T08:37:18","modified_gmt":"2026-04-07T15:37:18","slug":"agile-when-the-contract-says-no","status":"publish","type":"aa_event_session","link":"https:\/\/agilealliance.org\/resources\/sessions\/agile-when-the-contract-says-no\/","title":{"rendered":"Agile When the Contract Says \u201cNo\u201d"},"content":{"rendered":"\n

The following is an AI summary of the event.<\/em><\/p>\n\n\n\n

You can download the slide deck for this presentation at the bottom of the summary.<\/strong><\/p>\n\n\n\n

Overview<\/h2>\n\n\n\n

This session featured a real-world case study from Maria Ghlonti on delivering an AI product under tight constraints. The talk focused on how Agile practices were adapted in a high-pressure environment with fixed scope, budget, and timeline, and what actually made the project succeed.<\/p>\n\n\n\n

Project Context and Constraints<\/h2>\n\n\n\n

Maria described a project for a Berlin-based client with strict limitations:<\/p>\n\n\n\n

    \n
  • 3-month delivery window<\/li>\n\n\n\n
  • Fixed budget from an innovation fund that could not roll over<\/li>\n\n\n\n
  • Limited flexibility in contract terms<\/li>\n\n\n\n
  • Ambitious expectations despite constraints<\/li>\n<\/ul>\n\n\n\n

    Despite this setup, the project achieved a 5\/5 customer satisfaction score<\/strong>, which framed the discussion around what worked in practice.<\/p>\n\n\n\n

    Investing in People First<\/h2>\n\n\n\n

    With little room to adjust scope or budget, the main lever was team composition:<\/p>\n\n\n\n

      \n
    • Focus on bringing in highly capable, adaptable people<\/li>\n\n\n\n
    • Prioritize mindset and collaboration over just technical expertise<\/li>\n\n\n\n
    • Treat the team as the primary driver of success in a complex AI environment<\/li>\n<\/ul>\n\n\n\n

      Maria operated formally as a project manager but applied a Scrum Master mindset<\/strong>, emphasizing transparency, teamwork, and process over hierarchy.<\/p>\n\n\n\n

      Defining a Clear Business Goal<\/h2>\n\n\n\n

      A major early investment was aligning stakeholders around a single, clear objective:<\/p>\n\n\n\n

        \n
      • Move from ~4% of cases evaluated manually to 100% evaluated by AI with recommendations<\/strong><\/li>\n\n\n\n
      • Spend significant time (1 to 2 weeks) in workshops to align expectations<\/li>\n<\/ul>\n\n\n\n

        This clarity helped guide prioritization and trade-offs throughout the project.<\/p>\n\n\n\n

        Identifying the Real User and Decision-Maker<\/h2>\n\n\n\n

        Initially, the \u201cclient\u201d was assumed to be a platform team. That turned out to be incorrect.<\/p>\n\n\n\n

          \n
        • The platform team was a supporting function, not the end user<\/li>\n\n\n\n
        • The actual user and decision-maker had to be identified and engaged<\/li>\n\n\n\n
        • A true product owner role was established with clear responsibilities<\/li>\n<\/ul>\n\n\n\n

          This shift improved decision-making speed and reduced misalignment.<\/p>\n\n\n\n

          Managing the Client Relationship<\/h2>\n\n\n\n

          Maria highlighted a common issue in service-provider setups:<\/p>\n\n\n\n

            \n
          • Teams tend to over-accommodate clients to avoid conflict<\/li>\n\n\n\n
          • This can lead to poor decisions and weak outcomes<\/li>\n<\/ul>\n\n\n\n

            Instead, she:<\/p>\n\n\n\n

              \n
            • Set clear expectations with the product owner<\/li>\n\n\n\n
            • Positioned them as part of the team, not above it<\/li>\n\n\n\n
            • Encouraged shared decision-making and accountability<\/li>\n<\/ul>\n\n\n\n

              Agile Execution and Rapid Adaptation<\/h2>\n\n\n\n

              The team adapted its working model as complexity increased:<\/p>\n\n\n\n

                \n
              • Started with 2-week sprints<\/li>\n\n\n\n
              • Moved to weekly sprints<\/li>\n\n\n\n
              • Eventually shifted to daily iterations and feedback loops<\/strong><\/li>\n<\/ul>\n\n\n\n

                Key practices:<\/p>\n\n\n\n

                  \n
                • Daily review of AI outputs<\/li>\n\n\n\n
                • Continuous fine-tuning<\/li>\n\n\n\n
                • Nightly batch runs to generate new results<\/li>\n\n\n\n
                • Frequent replanning based on real feedback<\/li>\n<\/ul>\n\n\n\n

                  This created a tight learning loop between users and the AI system.<\/p>\n\n\n\n

                  Using AI as a Support Tool<\/h2>\n\n\n\n

                  AI was used in limited, practical ways:<\/p>\n\n\n\n

                    \n
                  • Refining backlog items and acceptance criteria<\/li>\n\n\n\n
                  • Generating additional ideas or options<\/li>\n<\/ul>\n\n\n\n

                    However:<\/p>\n\n\n\n

                      \n
                    • The backlog remained a human-driven thinking tool<\/li>\n\n\n\n
                    • AI supported decisions but did not replace them<\/li>\n<\/ul>\n\n\n\n

                      Handling Data and Infrastructure Limitations<\/h2>\n\n\n\n

                      A major challenge was poor data and testing infrastructure:<\/p>\n\n\n\n

                        \n
                      • \u201cGarbage in, garbage out\u201d directly impacted results<\/li>\n\n\n\n
                      • Traditional testing environments were not reliable<\/li>\n<\/ul>\n\n\n\n

                        Workarounds included:<\/p>\n\n\n\n

                          \n
                        • Using pre-production and production environments for testing<\/li>\n\n\n\n
                        • Running controlled, limited-access \u201cbeta\u201d usage in production<\/li>\n<\/ul>\n\n\n\n

                          This carried risk but allowed progress under constraints.<\/p>\n\n\n\n

                          Key Takeaways<\/h2>\n\n\n\n
                            \n
                          • Clarity beats complexity<\/strong>: A simple, shared goal enabled faster decisions<\/li>\n\n\n\n
                          • People matter most<\/strong>: Strong team dynamics outweighed process limitations<\/li>\n\n\n\n
                          • Real users drive success<\/strong>: Identifying the correct product owner was critical<\/li>\n\n\n\n
                          • Agility must increase with AI<\/strong>: Faster feedback loops were necessary, not optional<\/li>\n\n\n\n
                          • AI exposes weak systems<\/strong>: Data quality and infrastructure issues become immediate blockers<\/li>\n\n\n\n
                          • Adaptation is constant<\/strong>: The team changed not just the product, but how they worked<\/li>\n<\/ul>\n\n\n\n

                            Bottom Line<\/h2>\n\n\n\n

                            This case shows that AI product delivery is less about tools and more about alignment, feedback speed, and decision clarity<\/strong>. Under pressure, the team succeeded by tightening collaboration, shortening feedback cycles, and keeping ownership with people rather than the technology.<\/p>\n","protected":false},"excerpt":{"rendered":"

                            A real-world webinar case study on using Agile practices to deliver an AI product under tight timelines, fixed budgets, and constant change.<\/p>\n","protected":false},"author":8033092,"featured_media":8106831,"parent":0,"menu_order":0,"comment_status":"open","ping_status":"closed","template":"","categories":[883,908],"tags":[2197],"event_session_cat":[],"session_aud_level":[],"event_session_type":[],"content_source":[1697],"event_session_tags":[],"class_list":["post-8106810","aa_event_session","type-aa_event_session","status-publish","has-post-thumbnail","hentry","category-mindset","category-process","tag-free","content_source-agile-minicon"],"acf":[],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/aa_event_session\/8106810","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/aa_event_session"}],"about":[{"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/types\/aa_event_session"}],"author":[{"embeddable":true,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/users\/8033092"}],"replies":[{"embeddable":true,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/comments?post=8106810"}],"version-history":[{"count":4,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/aa_event_session\/8106810\/revisions"}],"predecessor-version":[{"id":8106830,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/aa_event_session\/8106810\/revisions\/8106830"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/media\/8106831"}],"wp:attachment":[{"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/media?parent=8106810"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/categories?post=8106810"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/tags?post=8106810"},{"taxonomy":"event_session_cat","embeddable":true,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/event_session_cat?post=8106810"},{"taxonomy":"session_aud_level","embeddable":true,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/session_aud_level?post=8106810"},{"taxonomy":"event_session_type","embeddable":true,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/event_session_type?post=8106810"},{"taxonomy":"content_source","embeddable":true,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/content_source?post=8106810"},{"taxonomy":"event_session_tags","embeddable":true,"href":"https:\/\/agilealliance.org\/wp-json\/wp\/v2\/event_session_tags?post=8106810"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}