Large Language Models (LLMs) are a new paradigm to computing. It changes fundamentally how to we get the job done. Before we get ahead of ourselves, lets overlook the work of a Indian Transporter. They are primarily tasked with brokering a transaction between a truck and a shippers looking to move their cargo. It involves three different jobs to be done like booking, tracking and payment faciliatation for every transaction. Each type of job requires its own specific worflow. [[Understandings Systems at work|Workflows]] are a interlinked tasks with a repeatabble pattern that have to be executed to get the job done. LLMs enable creative solutioning with incredible compute power to recieve a step change increase in productivity (througput of executing jobs). In domains like transportation, change management is a [[Understandings Systems at work|flip side]] of product adoption. So, any internal workflow that enables a stakeholder to become more productive is hard to sell due to the hurdle of change management. Instead, we augement the work of transporter by enabling their jobs to be done with agentic workflows. ![[Transporter_Workflows.png]] Let us take the case of booking of truck for transporter. We have a `Stakeholder DB` with all stakeholders involved for a booking. A transporter tasked with mapping a cargo of the shipper with a truck has to execute 4 workflows(approximate) to get this job done. With LLMs we build an agent that help a transporter execute all 4 worflows without changing the way they work. It replicates the mode of input(talk or rather shout at the agent) but then structure prompt and executes all 4 workflows simulataneously to get the task of booking a truck a fraction of time. - Inform the shipper of the price band for transportation. - Check availability of truck from within their network and negotiate a price with truck owner. - Reference history of lane to prioritise who to call. - Execute this multiple times till a match. All these take anywhere between couple of minutes to multiple hours depending on different scenarios. If these workflows are executed in seconds and then the data from the execution can be structured and stored for future reference. We can build agents that enables transactions in never before manner that are also configurable to every tranporters need without any form of standardising. I really liked the way Sarah Tavel summarised about the purpose of AI based on LLMs, [sell work not software. ](https://www.sarahtavel.com/p/ai-startups-sell-work-not-software) Hope this answers your question! Yours truly, Vivek, **** #replytoquestion #logistics Date : 02-10-24