Agentic Workflows

Follow this topic Ask about this topic

Agentic workflows study how language-model agents plan and carry out multi-step tasks by calling tools, coordinating specialized agents, and maintaining state or artifacts. Systems range from fixed graphs and typed protocols with deterministic checks to dynamically planned workflows that use reusable skills, human approval, and execution feedback. Work includes scientific and enterprise automation, workflow generation and evaluation, and serving methods such as model routing, caching, scheduling, and parallel execution. A central challenge is balancing flexible autonomy and efficiency with dependable, auditable behavior, since successful outputs can conceal unsafe shortcuts, semantic errors, privacy leaks, or failures that compound across steps.

Latest papers

218 papers20 papers in the last four weeks, up 18% on the four weeks before. 0.1% of all new papers.

Efficient Serving for Dynamic Agent Workflows with Prediction-based KV-Cache Management
May 7, 2026Haoyu Zheng, Fangcheng Fu, Jia Wu +6LLM Inference Efficiency KV Caching
Wuhan University · Shanghai Jiao Tong University · Macquarie University +2
Beyond Task Success: Measuring Workflow Fidelity in LLM-Based Agentic Payment Systems
May 7, 2026Donghao Huang, Joon Kiat Chua, Zhaoxia WangFinancial Services Multi-Agent LLM Systems
School of Computing and Information Systems, Singapore Management University, 80 Stamford Rd, Singapore 178902, Singapore · Research and Development, Mastercard, 4250 Fairfax Dr, Arlington, VA 22203, USA
From Agent Loops to Deterministic Graphs: Execution Lineage for Reproducible AI-Native Work
May 7, 2026Josh Rosen, Seth RosenAgentic Workflows Incremental Computation
ThruWire, Inc.
On Time, Within Budget: Constraint-Driven Online Resource Allocation for Agentic Workflows
May 7, 2026Xinglin Wang, Zishen Liu, Shaoxiong Feng +9Stochastic Optimization Cost-Aware Inference
School of Computer Science, Beijing Institute of Technology · Xiaohongshu Inc
GeoDecider: A Coarse-to-Fine Agentic Workflow for Explainable Lithology Classification
May 5, 2026Jiahao Wang, Mingyue Cheng, Yitong Zhou +4Explainable Artificial Intelligence Agentic Workflows
State Key Laboratory of Cognitive Intelligence, University of Science and Technology of China, Hefei, China
AAFLOW: Scalable Patterns for Agentic AI Workflows
May 4, 2026Arup Kumar Sarker, Mills Staylor, Aymen Alsaadi +3LLM Agent Orchestration Agentic Workflows
Department of Computer Science, Biocomplexity Institute and Initiative University of Virginia Charlottesville, VA, USA · Department of Computer Science, Rutgers University New Brunswick, NJ, USA · Biocomplexity Institute and Initiative University of Virginia Charlottesville, VA, USA +1
MedScribe: Clinically Grounded CT Reporting through Agentic Workflows
May 3, 2026Giuseppe A. Orlando, Paolo Papotti, Maria A. Zuluaga +2Computed Tomography Healthcare
Inria Centre at Université Côte d’Azur Sophia Antipolis, France · EURECOM Biot, France · Dep. of Nuclear Medicine, Centre Antoine Lacassagne Nice, France Université Côte d’Azur, CNRS, Inserm, iBV
Effect-Transparent Governance for AI Workflow Architectures: Semantic Preservation, Expressive Minimality, and Decidability Boundaries
May 1, 2026Alan L. McCannAgentic Workflows Formal Verification
Mashin, Inc.
Building Persona-Based Agents On Demand: Tailoring Multi-Agent Workflows to User Needs
Apr 30, 2026Giuseppe Arbore, Andrea Sillano, Luigi De RussisMulti-Agent Orchestration Agentic Workflows
Politecnico di Torino, Italy
Trace-Level Analysis of Information Contamination in Multi-Agent Systems
Apr 30, 2026Anna Mazhar, Huzaifa Suri, Sainyam GalhotraAgentic Workflows AI Agent Monitoring
Cornell University Ithaca, NY, USA · University of Illinois Urbana-Champaign, IL, USA
Pragmos: A Process Agentic Modeling System
Apr 30, 2026Pedro-Aarón Hernández-Ávalos, Luciano García-BañuelosAgentic Workflows Business Process Modeling
Tecnologico de Monterrey, Mexico
FlowBot: Inducing LLM Workflows with Bilevel Optimization and Textual Gradients
Apr 29, 2026Hongyeon Yu, Young-Bum Kim, Yoon KimAgentic Workflows LLM Agent Workflow Optimization
Naver Search US · Massachusetts Institute of Technology
Towards Agentic Investigation of Security Alerts
Apr 28, 2026Even Eilertsen, Vasileios Mavroeidis, Gudmund GrovCybersecurity Agentic Workflows
University of Oslo · Norwegian Defence Research Establishment (FFI) · & University of Oslo
Plausible but Wrong: A case study on Agentic Failures in Astrophysical Workflows
Apr 28, 2026Shivam Rawat, Lucie FlekAgent Failure Analysis AI for Science
Bonn-Aachen International Center for Information Technology, University of Bonn, Germany; Lamarr Institute for Machine Learning and Artificial Intelligence, Germany.
Why Search When You Can Transfer? Amortized Agentic Workflow Design from Structural Priors
Apr 27, 2026Shiyi Du, Jiayuan Liu, Weihua Du +6Agentic Workflows LLM-Based Program Synthesis
Carnegie Mellon University · Foundations of Cooperative AI Lab (FOCAL) · University of Notre Dame
Agentic AI for Remote Sensing: Technical Challenges and Research Directions
Apr 27, 2026Muhammad Akhtar Munir, Muhammad Umer Sheikh, Akashah Shabbir +5Agentic Workflows Remote Sensing
Mohamed bin Zayed University of Artificial Intelligence, Abu Dhabi, UAE · Technical University of Munich, Munich, Germany · BIFOLD and Technische Universität Berlin, Berlin, Germany
AgentEval: DAG-Structured Step-Level Evaluation for Agentic Workflows with Error Propagation Tracking
Apr 26, 2026Dongxin Guo, Jikun Wu, Siu Ming YiuAgent Failure Analysis AI Agent Evaluation
The University of Hong Kong · Stellaris AI Limited
A Decoupled Human-in-the-Loop System for Controlled Autonomy in Agentic Workflows
Apr 24, 2026Edward Cheng, Jeshua ChengAgentic Workflows Human-in-the-Loop AI
From Research Question to Scientific Workflow: Leveraging Agentic AI for Science Automation
Apr 23, 2026Bartosz Balis, Michal Orzechowski, Piotr Kica +2Agentic Workflows Scientific Workflow Automation
AGH University of Krakow, Krakow, Poland · Sano Centre for Computational Medicine, Krakow, Poland
GeoMind: An Agentic Workflow for Lithology Classification with Reasoned Tool Invocation
Apr 23, 2026Yitong Zhou, Mingyue Cheng, Jiahao Wang +2Evidence-Grounded Reasoning Agentic Workflows
State Key Laboratory of Cognitive Intelligence, University of Science and Technology of China, Hefei, China
Toward Reliable Design of LLM-Enabled Agentic Workflows: Optimizing Latency-Reliability-Cost Tradeoffs
Apr 21, 2026Ya-Ting Yang, Quanyan ZhuCost-Aware Inference Agentic Workflows
New York University Brooklyn, NY, USA
Chat2Workflow: A Benchmark for Generating Executable Visual Workflows with Natural Language
Apr 21, 2026Yi Zhong, Buqiang Xu, Yijun Wang +4Agentic Workflows Business Process Automation
Zhejiang University · Tencent
Aether: Network Validation Using Agentic AI and Digital Twin
Apr 20, 2026Jordan Auge, Sam Betts, Giovanna Carofiglio +3Agentic Workflows Digital Twins
Cisco Systems · Cisco Systems, Paris, France and London, UK. · Swisscom +1
The Semi-Executable Stack: Agentic Software Engineering and the Expanding Scope of SE
Apr 16, 2026Robert Feldt, Per Lenberg, Julian Frattini +1Software Engineering Software Engineering Agents
Chalmers University of Technology, Gothenburg, Sweden · Volvo Group, Gothenburg, Sweden
Scepsy: Serving Agentic Workflows Using Aggregate LLM Pipelines
Apr 16, 2026Otto White, Marcel Wagenländer, Britannio Jarrett +7LLM Inference LLM Inference Scheduling
Imperial College London · Independent Researcher
Agent-Aided Design for Dynamic CAD Models
Apr 16, 2026Mitch Adler, Matthew Russo, Michael CafarellaParametric CAD Modeling Agentic Workflows
Independent Researcher San Francisco, CA, USA · MIT Cambridge, MA, USA
SciFi: A Safe, Lightweight, User-Friendly, and Fully Autonomous Agentic AI Workflow for Scientific Applications
Apr 14, 2026Qibin Liu, Julia GonskiAgentic Workflows Scientific Workflow Automation
SLAC National Accelerator Laboratory, 2575 Sand Hill Rd, Menlo Park CA 94025 USA
From Intent to AI Pipelines: A Controlled Agentic Framework for Non-AI Expert Scientists
Apr 10, 2026Hyacinth Ali, Jessie Galasso-Carbonnel, Houari SahraouiAI-Assisted Scientific Research Agentic Workflows
University of Montreal, Montreal, Canada · McGill University, Montreal, Canada
VineLM: Trie-Based Fine-Grained Control for Agentic Workflows
Apr 9, 2026Nikos Pagonas, Matthew Lou, Tianyi Peng +2LLM Agent Workflow Optimization Agentic Workflows
Columbia University
Automotive Engineering-Centric Agentic AI Workflow Framework
Apr 9, 2026Tong Duy Son, Zhihao Liu, Piero Brigida +4Agentic Workflows Workflow Automation
The authors are with Siemens Digital Industries Software, Leuven, Belgium.

May 7, 2026J/K move · Enter open · S save

Efficient Serving for Dynamic Agent Workflows with Prediction-based KV-Cache Management

Haoyu Zheng, Fangcheng Fu, Jia Wu, Binhang Yuan+5

Wuhan University · Shanghai Jiao Tong University · Macquarie University+2

LLM Inference Efficiency KV Caching Agentic Workflows KV-Cache Management LLM Agent Workflow Optimization KV-Cache Reuse

Open paper AskPDF arXiv

LLM-based workflows compose specialized agents to execute complex tasks, and these agents usually share substantial context, allowing KV-Cache reuse to save computation. Existing approaches either manage KV-Cache at agent level and fail to exploit the reuse opportunities within workflows, or manage cache at the workflow level but assume that each workflow calls a static sequence of agents. However, practical workflows are typically dynamic, where the sequence of invoked agents and thus induced cache reuse opportunities depend on the context of each task. To serve such dynamic workflows efficiently, we build a system dubbed PBKV (Prediction-Based KV-Cache Management). For each workflow, PBKV predicts the agent invocations in several future steps by fusing the guidance from historical workflows and context of the target workflow. Based on the predictions, PBKV estimates the reuse potential of cache entries and keeps the high-potential entries in GPU memory. To be robust to prediction errors, PBKV utilizes the predictions conservatively during both cache eviction and prefetching. Experiments on three workflow benchmarks show that PBKV achieves up to

1.85\times

speedup over LRU on dynamic workflows, and up to

1.26\times

speedup over the SOTA baseline KVFlow on the static workflow.