Hasil untuk "Business"

Sumanth Balaji, Piyush Mishra, Aashraya Sachdeva et al.

Traditional customer support systems, such as Interactive Voice Response (IVR), rely on rigid scripts and lack the flexibility required for handling complex, policy-driven tasks. While large language model (LLM) agents offer a promising alternative, evaluating their ability to act in accordance with business rules and real-world support workflows remains an open challenge. Existing benchmarks primarily focus on tool usage or task completion, overlooking an agent's capacity to adhere to multi-step policies, navigate task dependencies, and remain robust to unpredictable user or environment behavior. In this work, we introduce JourneyBench, a benchmark designed to assess policy-aware agents in customer support. JourneyBench leverages graph representations to generate diverse, realistic support scenarios and proposes the User Journey Coverage Score, a novel metric to measure policy adherence. We evaluate multiple state-of-the-art LLMs using two agent designs: a Static-Prompt Agent (SPA) and a Dynamic-Prompt Agent (DPA) that explicitly models policy control. Across 703 conversations in three domains, we show that DPA significantly boosts policy adherence, even allowing smaller models like GPT-4o-mini to outperform more capable ones like GPT-4o. Our findings demonstrate the importance of structured orchestration and establish JourneyBench as a critical resource to advance AI-driven customer support beyond IVR-era limitations.

Jiazhu Xie, Bowen Li, Heyu Fu et al.

Large Language Model (LLM)-based question-answering systems offer significant potential for automating customer support and internal knowledge access in small businesses, yet their practical deployment remains challenging due to infrastructure costs, engineering complexity, and security risks, particularly in retrieval-augmented generation (RAG)-based settings. This paper presents an industry case study of an open-source, multi-tenant platform that enables small businesses to deploy customised LLM-based support chatbots via a no-code workflow. The platform is built on distributed, lightweight k3s clusters spanning heterogeneous, low-cost machines and interconnected through an encrypted overlay network, enabling cost-efficient resource pooling while enforcing container-based isolation and per-tenant data access controls. In addition, the platform integrates practical, platform-level defences against prompt injection attacks in RAG-based chatbots, translating insights from recent prompt injection research into deployable security mechanisms without requiring model retraining or enterprise-scale infrastructure. We evaluate the proposed platform through a real-world e-commerce deployment, demonstrating that secure and efficient LLM-based chatbot services can be achieved under realistic cost, operational, and security constraints faced by small businesses.

Nordine Benkeltoum

Background. Business Simulation Games (BSG) are widely used to foster experiential learning in complex managerial and organisational contexts by exposing students to decision-making under uncertainty. In parallel, Artificial Intelligence (AI) is increasingly integrated into higher education to support learning activities. However, despite growing interest of AI in education, its specific role in BSG and its implications for knowledge creation processes remain under-theorised. Intervention. This paper reports on the integration of generative AI tools into a BSG designed for engineering students. AI was embedded as a support mechanism during the simulation to assist students in analysing events, reformulating information, and generating decision-relevant insights, while instructors retained responsibility for supervision, debriefing, and complex issues. Methods. Adopting a qualitative experience-report approach, the study draws on the SECI model (Socialisation, Externalisation, Combination, Internalisation) as an analytical framework to examine how students and instructors interacted with AI during the simulation and how different forms of knowledge were mobilised and developed. Results. The findings indicate that AI primarily supports the Combination phase of the SECI model by facilitating the rapid synthesis, reformulation, and contextualisation of explicit knowledge. In contrast, the processes of Socialisation, Externalisation, and Internalisation remained largely dependent on peer interaction, individual reflection, and instructor guidance. Discussion. The results suggest a functional boundary in human-AI collaboration within simulation-based learning. AI acts as a cognitive enhancer that improves responsiveness and access to explicit knowledge, but it does not replace the pedagogical role of instructors in supporting the development of tacit knowledge, competencies, and phronesis. Conclusion. Integrating AI into BSG can enhance learning efficiency and engagement, but effective experiential learning continues to rely on active human supervision. Future research should investigate instructional designs that better support tacit knowledge acquisition in AI-assisted simulations.

Kartikey Singh Bhandari, Manav Ganesh, Yashwant Viswanathan et al.

Customer reviews contain detailed, domain specific signals about service failures and user expectations, but converting this unstructured feedback into actionable business decisions remains difficult. We study review-to-action generation: producing concrete, implementable recommendations grounded in review text. We propose a modular two-LLM framework in which an Issue model extracts salient issues and assigns coarse themes, and an Advice model generates targeted operational fixes conditioned on the extracted issue representation. To enable specialization without expensive full fine-tuning, we adapt the Advice model using a mixture of LoRA experts strategy: multiple low-rank adapters are trained and a lightweight gating mechanism performs token-level expert mixing at inference, combining complementary expertise across issue types. We construct synthetic review-issue-advice triples from Yelp reviews (airlines and restaurants) to supervise training, and evaluate recommendations using an eight dimension operational rubric spanning actionability, specificity, feasibility, expected impact, novelty, non-redundancy, bias, and clarity. Across both domains, our approach consistently outperforms prompting-only and single-adapter baselines, yielding higher actionability and specificity while retaining favorable efficiency-quality trade-offs.

Debarati Das, Khanh Chi Le, Ritik Sachin Parkar et al.

Legal practitioners, particularly those early in their careers, face complex, high-stakes tasks that require adaptive, context-sensitive reasoning. While AI holds promise in supporting legal work, current datasets and models are narrowly focused on isolated subtasks and fail to capture the end-to-end decision-making required in real-world practice. To address this gap, we introduce LawFlow, a dataset of complete end-to-end legal workflows collected from trained law students, grounded in real-world business entity formation scenarios. Unlike prior datasets focused on input-output pairs or linear chains of thought, LawFlow captures dynamic, modular, and iterative reasoning processes that reflect the ambiguity, revision, and client-adaptive strategies of legal practice. Using LawFlow, we compare human and LLM-generated workflows, revealing systematic differences in structure, reasoning flexibility, and plan execution. Human workflows tend to be modular and adaptive, while LLM workflows are more sequential, exhaustive, and less sensitive to downstream implications. Our findings also suggest that legal professionals prefer AI to carry out supportive roles, such as brainstorming, identifying blind spots, and surfacing alternatives, rather than executing complex workflows end-to-end. Our results highlight both the current limitations of LLMs in supporting complex legal workflows and opportunities for developing more collaborative, reasoning-aware legal AI systems. All data and code are available on our project page (https://minnesotanlp.github.io/LawFlow-website/).

Yaw Osei Adjei, Frederick Ayivor

Business Email Compromise (BEC) is a high-impact social engineering threat with extreme operational asymmetry: false negatives can trigger large financial losses, while false positives primarily incur investigation and delay costs. This paper compares two BEC detection paradigms under a cost-sensitive decision framework: (i) a semantic transformer approach (DistilBERT) for contextual language understanding, and (ii) a forensic psycholinguistic approach (CatBoost) using engineered linguistic and structural cues. We evaluate both on a hybrid dataset (N = 7,990) combining legitimate corporate email and AI-synthesised adversarial fraud generated across 30 BEC taxonomies, including character-level Unicode obfuscations. We add classical baselines (TF-IDF+LogReg and character n-gram+Linear SVM), an ablation study for the Smiling Assassin Score, and a homoglyph-map sensitivity analysis. DistilBERT achieves AUC = 1.0000 and F1 = 0.9981 at 7.403 ms per email on GPU; CatBoost achieves AUC = 0.9860 and F1 = 0.9382 at 0.855 ms on CPU. A three-way cost-sensitive decision policy (auto-allow, auto-block, manual review) optimises expected financial loss under a 1:5,167 false-negative-to-false-positive cost ratio.

Sonia Deshmukh, Shikha Gupta, Naveen Kumar

Process discovery approaches analyze the business data to automatically uncover structured information, known as a process model. The quality of a process model is measured using quality dimensions -- completeness (replay fitness), preciseness, simplicity, and generalization. Traditional process discovery algorithms usually output a single process model. A single model may not accurately capture the observed behavior and overfit the training data. We have formed the process discovery problem in a multi-objective framework that yields several candidate solutions for the end user who can pick a suitable model based on the local environmental constraints (possibly varying). We consider the Binary Differential Evolution approach in a multi-objective framework for the task of process discovery. The proposed method employs dichotomous crossover/mutation operators. The parameters are tuned using Grey relational analysis combined with the Taguchi approach. {We have compared the proposed approach with the well-known single-objective algorithms and state-of-the-art multi-objective evolutionary algorithm -- Non-dominated Sorting Genetic Algorithm (NSGA-II).} Additional comparison via computing a weighted average of the quality dimensions is also undertaken. Results show that the proposed algorithm is computationally efficient and produces diversified candidate solutions that score high on the fitness functions. It is shown that the process models generated by the proposed approach are superior to or at least as good as those generated by the state-of-the-art algorithms.

Michael Wornow, Avanika Narayan, Ben Viggiano et al.

Existing ML benchmarks lack the depth and diversity of annotations needed for evaluating models on business process management (BPM) tasks. BPM is the practice of documenting, measuring, improving, and automating enterprise workflows. However, research has focused almost exclusively on one task - full end-to-end automation using agents based on multimodal foundation models (FMs) like GPT-4. This focus on automation ignores the reality of how most BPM tools are applied today - simply documenting the relevant workflow takes 60% of the time of the typical process optimization project. To address this gap we present WONDERBREAD, the first benchmark for evaluating multimodal FMs on BPM tasks beyond automation. Our contributions are: (1) a dataset containing 2928 documented workflow demonstrations; (2) 6 novel BPM tasks sourced from real-world applications ranging from workflow documentation to knowledge transfer to process improvement; and (3) an automated evaluation harness. Our benchmark shows that while state-of-the-art FMs can automatically generate documentation (e.g. recalling 88% of the steps taken in a video demonstration of a workflow), they struggle to re-apply that knowledge towards finer-grained validation of workflow completion (F1 < 0.3). We hope WONDERBREAD encourages the development of more "human-centered" AI tooling for enterprise applications and furthers the exploration of multimodal FMs for the broader universe of BPM tasks. We publish our dataset and experiments here: https://github.com/HazyResearch/wonderbread

Dirk Fahland, Fabiana Fournier, Lior Limonad et al.

Large Language Models (LLMs) are trained on a vast amount of text to interpret and generate human-like textual content. They are becoming a vital vehicle in realizing the vision of the autonomous enterprise, with organizations today actively adopting LLMs to automate many aspects of their operations. LLMs are likely to play a prominent role in future AI-augmented business process management systems, catering functionalities across all system lifecycle stages. One such system's functionality is Situation-Aware eXplainability (SAX), which relates to generating causally sound and human-interpretable explanations. In this paper, we present the SAX4BPM framework developed to generate SAX explanations. The SAX4BPM suite consists of a set of services and a central knowledge repository. The functionality of these services is to elicit the various knowledge ingredients that underlie SAX explanations. A key innovative component among these ingredients is the causal process execution view. In this work, we integrate the framework with an LLM to leverage its power to synthesize the various input ingredients for the sake of improved SAX explanations. Since the use of LLMs for SAX is also accompanied by a certain degree of doubt related to its capacity to adequately fulfill SAX along with its tendency for hallucination and lack of inherent capacity to reason, we pursued a methodological evaluation of the perceived quality of the generated explanations. We developed a designated scale and conducted a rigorous user study. Our findings show that the input presented to the LLMs aided with the guard-railing of its performance, yielding SAX explanations having better-perceived fidelity. This improvement is moderated by the perception of trust and curiosity. More so, this improvement comes at the cost of the perceived interpretability of the explanation.

Geoffrey Goodell

This document constitutes a response to a Consultation Paper published by the Bank of England and HM Treasury, "The digital pound: a new form of money for households and businesses?", the latest document in a series that includes "Central Bank Digital Currency: opportunities, challenges and design" in 2020 and "New forms of digital money" in 2021. The Consultation Paper concerns the adoption of central bank digital currency (CBDC) for retail use in the United Kingdom by the Bank of England. We shall address the consultation questions directly in the third section of this document.

Guilherme Toledo, Yoni Zohar, Clark Barrett

This work is a part of an ongoing effort to understand the relationships between properties used in theory combination. We here focus on including two properties that are related to shiny theories: the finite model property and stable finiteness. For any combination of properties, we consider the question of whether there exists a theory that exhibits it. When there is, we provide an example with the simplest possible signature. One particular class of interest includes theories with the finite model property that are not finitely witnessable. To construct such theories, we utilize the Busy Beaver function.

Alex Felice, Lucija Rakocevic, Leen Peters et al.

Renewable energy communities (RECs) are prominent initiatives to provide end consumers an active role in the energy sector, raise awareness on the importance of renewable energy (RE) technologies and increase their share in the energy system thus reducing greenhouse gas emissions. The economic viability of RECs though, depends on multiple interdependent factors that require careful examination for each individual context. This study aims at investigating the impact of electricity tariffs, ratio of electrification of heating and transportation sectors, prices of RE technologies and storage systems, and internal electricity exchange prices on the annual cost for electricity provision of a REC. A mixed-integer linear model is developed to minimize energy provision costs for a representative REC in Flanders, Belgium. The results indicate that RECs have the potential to reduce these costs by 10 to 26% compared to business-as-usual. This cost reduction depends on the type of electricity tariffs and the level of uptake of flexible assets such as heat pumps and electric vehicles. The shift towards a higher power component in the electricity tariff makes electricity storage systems more attractive, which leads to higher electricity self-consumption. The introduction of flexible assets adds the possibility to shift demand when tariffs are lower and makes larger sizes of photovoltaic systems economically viable due to the increase in the total electricity demand. However, RECs cost reduction compared to individual smart-homes amounts to only 4% - 6% in the best cases. Uncertainties stemming from the regulation and the costs of setting up a REC may reduce the estimated benefits.

Usman Rafiq

Analytics plays a crucial role in the data-informed decision-making processes of modern businesses. Unlike established software companies, software startups are not seen utilizing the potential of analytics even though a startup process should be primarily data-driven. There has been little understanding in the literature about analytics for software startups. This study set out to address the knowledge gap by exploring how analytics is understood in the context of software startups. To this end, we collected the qualitative data of three analytics platforms that are mostly used by startups from multiple sources. We covered platform documentation as well as experience reports of the software startups using these platforms. The data was analyzed using content analysis techniques. Four high-level concepts were identified that encapsulate the real understanding of software startups on analytics, including instrumentation of analytics, experimentation, diagnostic analysis, and getting insights. The first concept describes how startups set up analytics and the latter three illustrate the usage scenarios of analytics. This study is the first step toward understanding analytics in the software startup context. The identified concepts can guide further investigation of analytics in this context. It also provides some insights for software startups to set up analytics for data-informed decisions. Given the limitation of the data used in the study, the immediate next step is to ground as well as validate the acquired understanding using the primary data, by directly interacting with software startups.

Maxime Elkael, Massinissa Ait Aba, Andrea Araldo et al.

In this article, we consider the Virtual Network Embedding (VNE) problem for 5G networks slicing. This problem requires to allocate multiple Virtual Networks (VN) on a substrate virtualized physical network while maximizing among others, resource utilization, maximum number of placed VNs and network operator's benefit. We solve the online version of the problem where slices arrive over time. Inspired by the Nested Rollout Policy Adaptation (NRPA) algorithm, a variant of the well known Monte Carlo Tree Search (MCTS) that learns how to perform good simulations over time, we propose a new algorithm that we call Neighborhood Enhanced Policy Adaptation (NEPA). The key feature of our algorithm is to observe NRPA cannot exploit knowledge acquired in one branch of the state tree for another one which starts differently. NEPA learns by combining NRPA with Neighbordhood Search in a frugal manner which improves only promising solutions while keeping the running time low. We call this technique a monkey business because it comes down to jumping from one interesting branch to the other, similar to how monkeys jump from tree to tree instead of going down everytime. NEPA achieves better results in terms of acceptance ratio and revenue-to-cost ratio compared to other state-of-the-art algorithms, both on real and synthetic topologies.

Manuel Camargo, Marlon Dumas, Oscar González-Rojas

Business process simulation is a well-known approach to estimate the impact of changes to a process with respect to time and cost measures -- a practice known as what-if process analysis. The usefulness of such estimations hinges on the accuracy of the underlying simulation model. Data-Driven Simulation (DDS) methods leverage process mining techniques to learn process simulation models from event logs. Empirical studies have shown that, while DDS models adequately capture the observed sequences of activities and their frequencies, they fail to accurately capture the temporal dynamics of real-life processes. In contrast, generative Deep Learning (DL) models are better able to capture such temporal dynamics. The drawback of DL models is that users cannot alter them for what-if analysis due to their black-box nature. This paper presents a hybrid approach to learn process simulation models from event logs wherein a (stochastic) process model is extracted via DDS techniques, and then combined with a DL model to generate timestamped event sequences. An experimental evaluation shows that the resulting hybrid simulation models match the temporal accuracy of pure DL models, while partially retaining the what-if analysis capability of DDS approaches.

Emanuele Fabbiani, Andrea Marziali, Giuseppe De Nicolao

Energy companies need efficient procedures to perform market calibration of stochastic models for commodities. If the Black framework is chosen for option pricing, the bottleneck of the market calibration is the computation of the variance of the asset. Energy commodities are commonly represented by multi-factor linear models, whose variance obeys a matrix Lyapunov differential equation. In this paper, analytical and numerical methods to derive the variance are discussed: the Lyapunov approach is shown to be more straightforward than ad-hoc derivations found in the literature and can be readily extended to higher-dimensional models. A case study is presented, where the variance of a two-factor mean-reverting model is embedded into the Black formulae and the model parameters are calibrated against listed options. The analytical and numerical method are compared, showing that the former makes the calibration 14 times faster. A Python implementation of the proposed methods is available as open-source software on GitHub.

José Apolinário Teixeira, Sami Hyrynsalmi

Much research that analyzes the evolution of a software ecosystem is confined to its own boundaries. Evidence shows, however, that software ecosystems co-evolve independently with other software ecosystems. In other words, understanding the evolution of a software ecosystem requires an especially astute awareness of its competitive landscape and much consideration for other software ecosystems in related markets. A software ecosystem does not evolve in insulation but with other software ecosystems. In this research, we analyzed the OpenStack software ecosystem with a focal perspective that attempted to understand its evolution as a function of other software ecosystems. We attempted to understand and explain the evolution of OpenStack in relation to other software ecosystems in the cloud computing market. Our findings add to theoretical knowledge in software ecosystems by identifying and discussing seven different mechanisms by which software ecosystems mutually influence each other: sedimentation and embeddedness of business relationships, strategic management of the portfolio of business relationships, firms values and reputation as a partner, core technological architecture, design of the APIs, competitive replication of functionality and multi-homing. Research addressing the evolution of software ecosystem should, therefore, acknowledge that software ecosystems entangle with other software ecosystems in multiple ways, even with competing ones. A rigorous analysis of the evolution of a software ecosystem should not be solely confined to its inner boundaries.

Jessica Chang, Samir Khuller, Koyel Mukherjee

We consider fundamental scheduling problems motivated by energy issues. In this framework, we are given a set of jobs, each with a release time, deadline and required processing length. The jobs need to be scheduled on a machine so that at most g jobs are active at any given time. The duration for which a machine is active (i.e., "on") is referred to as its active time. The goal is to find a feasible schedule for all jobs, minimizing the total active time. When preemption is allowed at integer time points, we show that a minimal feasible schedule already yields a 3-approximation (and this bound is tight) and we further improve this to a 2-approximation via LP rounding techniques. Our second contribution is for the non-preemptive version of this problem. However, since even asking if a feasible schedule on one machine exists is NP-hard, we allow for an unbounded number of virtual machines, each having capacity of g. This problem is known as the busy time problem in the literature and a 4-approximation is known for this problem. We develop a new combinatorial algorithm that gives a 3-approximation. Furthermore, we consider the preemptive busy time problem, giving a simple and exact greedy algorithm when unbounded parallelism is allowed, i.e., g is unbounded. For arbitrary g, this yields an algorithm that is 2-approximate.

Alexander Iksanov, Wissem Jedidi, Fethi Bouzeffour

We discuss weak convergence of the number of busy servers in a $G/G/\infty$ queue in the $J_1$-topology on the Skorokhod space. We prove two functional limit theorems, with random and nonrandom centering, respectively, thereby solving two open problems stated in Mikosch and Resnick (2006}. A new integral representation for the limit Gaussian process is given.

Fatima Boumahdi, Rachid Chalal

This paper presents an architecture for the Project Management, which is defined using the concepts behind ServiceOriented and Decision Support System. The framework described, denominated as SoaDssPm, represents the following: a coherent solution to the problem of control Project Management the existing gap between the real execution of Project Management by describing the business process and relationships required by a SOA solution, and its objectives representation, in which the decisional aspects determine the final shape of the system, providing decision support to the identified business processes and constraints.