Harnesswalk: why do people fine-tune at all?
The first post was about seeing inside an agent harness: the guardrails, context, memory, policy and approvals that sit around a model. This one is about the part I had treated as fixed. The model.
The first post was about seeing inside an agent harness: the guardrails, context, memory, policy and approvals that sit around a model. This one is about the part I had treated as fixed. The model.
An AI agent is more than a model call and a tool loop. Around that loop sit input guardrails, a context builder, memory, policy and approvals, output guardrails, tracing, evaluations and releases. You can read about those pieces in any architecture diagram. You rarely get to watch them work.
We’re now moving on to the third phase of the project, which involves the integration of an onboard 4G/LTE modem to enable remote outdoor mobility. Additionally, a new set of sensors has been introduced for telemetry transmission via AWS IoT Core using MQTT.
This is the second part of the project that tackles on sending a peer-to-peer video stream from the rover’s onboard camera into our frontend mobile application using WebRTC and AWS Kinesis Video Stream (KVS) as the Signaling Server.
This is the first part of the project which focuses on controlling the motors on a rover over the internet via MQTT. This article covers the hardware and software that make the whole system work.
When deploying an IoT project in a remote or offshore area, getting Internet connectivity (4G/LTE) is difficult. One solution for this is to communicate with satellites. This article will demonstrate how to talk with the Iridium 9602/3 module via the serial port using Nodejs.
This is a quick implementation of a File Uploader API that is deployed in AWS using API Gateway, Lambda, and S3. This API can upload any file type. We will be using Nodejs along with serverless framework to help us easily deploy services in AWS.
There are a bunch of ways to build a REST API, and this article here shows one of those ways. We will be building a serverless backend, so we won’t have to worry about setting up and maintaining the infrastructure, which could include load-balancing, horizontal scaling, security, etc. This implementation also includes log monitoring (through AWS CloudWatch).
The goal of this article is to create a very simple low level IoT application. This focuses more on how to build it from the ground up without using existing IoT Platform as service such AWS IoT, IBM Bluemix, Samsung or Artik.