Page 570 - AI for Good Innovate for Impact
P. 570

AI for Good Innovate for Impact



                      [1][2]. This loop—data ‐ persona adjustment ‐ strategy execution ‐ new data—allows the AI to
                      dynamically shift its persona and tactics to maximize the probability of a favorable outcome.
                      Integrated seamlessly with KT Commerce's platform, enhancing negotiation capabilities by
                      aligning real-time market prices with vendor responses [7].

                      (Short Version) AI personas in Nego-Wiz continually update by comparing live negotiation inputs
                      (e.g., partner’s last offer, market prices) stored in short-term memory against past outcomes
                      in long-term memory. A “Self-Reflection” agent re-evaluates which persona (cooperative,
                      competitive, neutral) best fits the evolving context. That persona then guides the “Actor” agent’s
                      responses, such as making a concession or holding firm. As each new offer or data point arrives,
                      the system re-scores and switches personas as needed, ensuring tactics adapt dynamically to
                      maximize success [4].

                      3.    Aggregates and analyzes real-time data from diverse sources across the organization,
                      delivering unbiased, organization-wide insights and recommendations that reduce individual
                      bias, support data-driven negotiation strategies, and enable consistent, optimal decision-
                      making for all stakeholders.

                      Expected Benefits:

                      •    The solution drastically reduces MD workload by automating price negotiation, allowing
                           MDs to focus on more strategic tasks.
                      •    Enhanced Negotiation Outcomes: Access to comprehensive, organization-wide insights
                           allows for more strategic and effective negotiation strategies, leading to improved pricing,
                           higher success rates, and greater overall business value.
                      •    Increased Efficiency and Productivity: Automating data analysis and insight generation
                           reduces manual effort, accelerates decision-making processes, and allows staff to focus
                           on higher-value tasks.
                      •    Greater Transparency and Collaboration: Organization-wide data integration fosters
                           transparency and enables collaborative negotiation, ensuring all relevant perspectives
                           are considered and supporting more equitable business outcomes.
                      •    Unbiased, Organization-Wide Decision-Making: By aggregating and analyzing real-time
                           data from diverse sources, the solution eliminates individual bias and delivers consistent,
                           data-driven insights, enabling all stakeholders to make optimal decisions that benefit the
                           entire organization.
                      •    Continuous Improvement: The AI system learns from every negotiation, continuously
                           refining its recommendations and strategies, which leads to ongoing optimization of
                           negotiation processes and outcomes across the organization.

                      As a result, KT Commerce has successfully automated and optimized price negotiations,
                      leading to a significant reduction in negotiation time, better alignment of product pricing,
                      and negotiation success rate.

                      Partners

                      KT Commerce provides of 60,000+ annual transaction datasets and AI negotiation PoC
                      verification.
                      Korea University HI-AI Laboratory provided advisory and support for the development of core
                      technologies in artificial intelligence and natural language processing.








                  534
   565   566   567   568   569   570   571   572   573   574   575