scheduling

Testing of LuMay Voice Agent in a healthcare clinic for 30 days showed improved response availability, patient willingness to use AI for routine scheduling, and reduced staff workload.

This paper analyzes the trade-off between mixed batching and exclusive batching for LLM inference, showing that the optimal choice depends on GPU memory bandwidth. It proposes a threshold-based hybrid scheduler that dynamically switches between the two methods, achieving up to 41.9% higher throughput on bandwidth-constrained GPUs.

This article advocates for using systemd timers over traditional cron jobs for scheduled tasks on Linux, highlighting their better integration, logging, and clearer syntax.

This paper presents a comprehensive survey and taxonomy of federated learning over human-body communication for on-body edge intelligence, including a scheduling vignette called BODYFED-HBC.

This paper presents the aircraft disassembly scheduling problem, a large-scale combinatorial optimization task involving thousands of tasks, precedence relations, balance constraints, and limited space. It proposes a Constraint Programming model and a MIP model tested on real operational instances with up to 1450 tasks.

This paper presents a hybrid approach combining dynamic programming and constraint programming to solve the Partial Shop Scheduling Problem, demonstrating the viability of integrating both paradigms despite not outperforming pure CP solvers.

Remind is a sophisticated calendar and alarm program with a scripting language, multiple output formats, and support for Hebrew calendar and multiple languages.

PALS is a power-aware runtime for LLM serving that treats GPU power caps as a controllable knob, jointly optimizing them with batch size to maximize energy efficiency while meeting throughput targets. The system improves energy efficiency by up to 26.3% and reduces QoS violations by 4x-7x under power constraints.

The Cal.com team has open sourced their entire scheduling platform as cal.diy under an MIT license, offering a free, self-hosted alternative to paid services like Calendly and SavvyCal.

Auto Posts is a tool for scheduling social media posts and Telegram messages.

The article discusses using Google's OR-Tools CP-SAT solver to optimize maintenance scheduling for cloud infrastructure at Akamai, addressing complex constraints like capacity and concurrency.

Espa is an AI executive assistant designed to help users manage their to-do lists, scheduling, and email.

ClawTick is a developer tool that enables cron job scheduling for AI agents with a single command and no infrastructure setup required.

Cal.com releases Cal.diy, a fully open-source MIT-licensed community fork with all enterprise features stripped out for personal self-hosting.

Toki 2.0 launches to automatically transform ideas into scheduled plans.