One Point of Interaction

DevOps Challenge: During issue resolution, site reliability engineers must access many different data sources across multiple windows. These interactions themselves can be a valuable source of expert knowledge that often goes undocumented. It can be hard to retrace the steps taken with disparate sources to reuse that deep dive analysis for solving similar issues in the future.

DagKnows Solution: Our browser-based Virtual Intelligent Collaborative Shell (VICS) works to centralize and capture all these important data acquisition interactions.

Unified View

Our unique VICS interface enables access to everything: machines, tools, knowledge base, and automation runbooks — all from one single window. Reduce visual complexity and streamline your workflows for greater efficiency. Executing commands on hosts is as simple as typing @hostname command. No need to create separate consoles for different machines.

Collaboration in VICS sessions

DagKnows let’s you create an independent troubleshooting sessions for each problem. Create a session and invite users for real-time collaborative debugging by sharing a URL to the session. Our advanced Role Based Access Control (RBAC) can be used to limit or grant the resource access privileges to the collaborating users.

Knowledge Management Made Easy

DevOps Challenge: After an engineer investigates and resolves issues, they are often required to write detailed knowledge articles ex-post-facto. At best this becomes an arduous task for them and at worst the task is ignored completely. As a result, knowledge bases are either missing crucial contextual details or are filled with poorly written articles which don’t serve their purpose.

DagKnows Solution: DagKnows VICS enables quick, in-line note taking. No context switching is required. No authoring tools need to be brought up. No lengthy knowledge articles need to be rewritten from memory.

Automatic Knowledge Creation

Type your troubleshooting commands to investigate the issue in VICS and DagKnows will automatically convert them into a well written knowledge nugget. Our AI engine figures our the intent behind your command and describes it English for documentation. It also converts your command into a reusable parameterized script. These knowledge nuggets with embedded scripts become the lego blocks for you to build more complex runbooks with very little efforts.

Easy Knowledge Authoring

You can also explicitly take notes and create knowledge nuggets with our customizable knowledge templates. These templates offer every feature you need to create a perfect knowledge articles with embedded images, videos, and edit rich text, and of course your curated scripts.

Intelligent Adaptive Search

Search for technical symptoms in your natural language and DagKnows NLP engine will find the most relevant runbooks and other resources to accelerate problem solving. We build a custom vocabulary for your particular environment and provide contextual search results when and where they are needed the most.

Structuring Knowledge for Reusability

DevOps Challenge: When users write knowledge articles or root cause analysis details in tickets, that knowledge isn’t always formatted or positioned well for reusability. Step-by-step guidance is the ideal for those who need to repeat the process of troubleshooting and remediation. Other knowledge authoring tools don’t provide this depth and structure for documenting algorithmic knowledge.

DagKnows Solution: DagKnows goes well beyond simple note taking. It brings an intelligent structure to your team’s cumulative knowledge. This makes it suitable for rapid investigation and remediation thereby paving the way for full or partial automation.

Deriving Causal Knowledge

One-line notes written in natural language are algorithmically mapped by DagKnows onto conceptual links between symptoms and causes. Directed Acyclic Graphs (DAG) are one of the ML systems working in the background to align your solution knowledge to symptoms and root-causes. Every node in these graphs is a problem statement and the children nodes are the causes of the problem. This structured knowledge is ideally suited for rapid traversal to the root-cause automatically.

One Click Root Cause Analysis

Data structures like DAGs make your root-cause analysis as easy as possible. With just one click, DagKnows can highlight the most likely root-cause so you know just what to focus on when troubleshooting. Discovering all the known causes or root causes of a symptom has never been easier than this.

Guided Resolution

DagKnows applies machine learning algorithms to compute the probability of each individual problem. It then guides you to a resolution by surfacing the next symptom you should focus on. You simply confirm yes/no on symptoms and let DagKnows assist you in minimizing your troubleshooting efforts.

Intelligent Runbooks

DevOps Challenge: During troubleshooting, users unknowingly just repeat the steps to first identify and then rule out the known issues. That wastes a lot of time which can saved if they could automate the diagnosis itself. Unfortunately the vanilla runbook tools don’t really allow them to capture the complex nuances of troubleshooting logic which requires branching to traverse to the root-cause.

DagKnows Solution: Our DAGs provide a simple yet powerful low-code framework for diagnosing problems which you can quickly execute to immediately rule out all the known issues captured previously in the DAG.

Diagnosing With Scripted Nodes

Every node in the DAG is a problem symptom. A simple script can be embedded in the node to detect the problem. Then execute the script to see if that node is a problem or not. It is just that easy.

Automated RCA

Any node in the troubleshooting DAG can be executed to trigger automatic execution for all child nodes underneath to identify where the problem originated. DagKnows algorithmically orders the execution of nodes so that the most likely root cause is detected first.

Remediation Actions

Each node also has a remediation section where you can embed script(s) to fix known issues. Either wait for manual input for custom remediation or optionally execute it automatically when the node shows a problem.