Agentic Service Management 101 – Getting started with AI agents for ITSM professionals on Nov 14, 2024. Register now ->

In This Article:

No items found.

Share Article

Atomicwork year in review: 2023

What we did in 2023, what's in store for 2024, and how we do it.

Let’s take a look back on a year filled with more AI than ever before (and take a sneak peek at what’s coming in 2024).

Where Atomicwork focused on innovation in 2023

  • We launched an employee assistant for Slack and Teams to help employees get instant answers to questions they might otherwise have to wait on IT/HR teams for a response.
  • We launched two knowledge integrations (Notion and SharePoint) and built support for document formats like docx, ppt, csv etc., so that teams could bring their knowledge to Atomicwork easily.
  • We built a request and incident management system to help IT and HR teams easily handle the questions that Atom couldn’t answer. without the employee having to log their ticket in a separate portal and keep track of it.
  • We launched workflows to help IT teams prioritize and route incoming requests, and setup approval flows. These include not just native actions but also actions in apps like Azure Active Directory, Microsoft Teams, Okta, Slack, Google Workspace, and Checkr. For example, an IT manager can setup a workflow to automatically add an employee to Azure AD groups based on requests.

Where we’re going in Q1 2024

  • We’re building insights to help teams figure out what knowledge they need to collate and areas employees need help in.
  • We’re bringing in asset management capabilities to help IT and HR teams keep track of assets and easily link issues to relevant assets.
  • We're spreading the conversational love to email as well by adding email support. IT and HR teams will be able to provide support via email through Atomicwork.
  • Mo’ workflows, mo’ actions. We’re building actions in Jira, Dropbox, Box, SharePoint, and so on.
  • Workflows will also expand beyond the realm of “if this, then that” to include conversational triggers. For example, an employee asking Atom to be added to a distribution list will trigger the DL workflow.

Atomicwork is on a mission to help businesses deliver great employee experiences and unlock productivity with a conversational ITSM solution. The only one with instant AI assistance, employee service delivery, and workflow automation.

This means that we have to reinvent how things are done while preserving the spirit of the philosophy that enables hundreds and thousands of service teams to provide world-class employee support without missing a beat.

Much like ITIL’s guiding principles, we also had to develop a set of guiding principles for the product team so everything we built further our ambition of bringing AI into the workplace in a way that created sustained value for IT teams and didn’t end up being just a thin software wrapper on the newest AI tech to come our way.

These guiding principles eventually became our ITSM manifesto.

To that end, we’ve focused on:

1. People-centricity

When we say people, we mean both employees and IT and HR teams. Our goal is to help IT and HR teams become more efficient while not sacrificing their goal of employee engagement.

IT and HR teams needed an Assistant that could engage employees without making them feel frustrated that they have to go through an IVR or a robotic chatbot to get help. Employees wouldn’t continue using a chatbot that made them feel like they’re not valued so we had to build an Assistant that could act like a human.

We decided that our AI Assistant, Atom would have to:

  • Integrate with Slack and Microsoft Teams where employees spend most of their time.
  • Converse with employees like a human would. This meant full sentences, a sense of humor and the ability to maintain context over the course of a conversation and remember previous inputs and responses. Just like a human would. Piece of cake, clearly.
        - A corollary of this is Atom recognizing and remembering an employee’s location, their role and department just like their HR-BP or IT manager would while answering a question.
  • Learn and update itself, like a human would, from conversations it has and what other people have in channels it’s a part of.
  • Learn on its own like a person would from the knowledge base. IT and HR teams shouldn’t have to spend a lot of time training it. Quick, easy to setup, deploy and add onto was our mantra. So we built Notion and SharePoint integrations along with the ability to upload PDFs, PPTs and CSV files for Atom to learn from. Atom also learns from document syncs every day to stay updated.
  • Help its teammates, just like a human would. We built features to assist agents with otherwise-tedious tasks like request summarization and spellcheck.

2. Context-based automation

Automation without context can be ineffective and even counterproductive because context provides the necessary information and understanding for meaningful decision-making. We wanted IT and HR teams to be able to truly delegate even tasks that require meaningful decision threads to Atomicwork.

We made sure that:

  • Autonomous actions are easy to set up. Siloed apps are a way of life for most IT and HR teams so our goal was to build a product that could serve as a bridge for the entire EX stack and make it easy for people to get more done. Whether it’s providing access to software (or revoking it), automating approval flows or running a security refresher course for employees, Atomicwork had to be end-to-end.
  • Workflows include actions from flow-of-work as well. For example, when incidents are created, Atom will kickstart workflows to notify stakeholders, create channels, and even generate an incident report making life easier for agents.
  • Workflows include decision paths so that IT managers and HR admins could build entire flows in Atomicwork and every scenario can be handled in some form or the other by Atomicwork.

3. Responsible AI

Ensuring responsible AI practices is at the core of our development philosophy.

  • We prioritize data security through SOC2 compliance, adhering to strict standards for customer data protection. Our AI models are equipped with guardrails, blacklisting topics like politics and religion to ensure ethical and unbiased responses. In preparation for GDPR, we’ve aligned our practices with regulatory requirements, emphasizing data protection and privacy.
  • Sensitive data is safeguarded using advanced techniques like data masking, reducing the risk of unauthorized access and breaches. Our responsible AI framework also ensures models don't train on customer data, prioritizing user confidentiality and trust.
  • Employees can check for biases and tone in AI-generated responses, promoting transparency and accountability.
  • Access controls, including segments and permissions, restrict sensitive information to individuals on a need-to-know basis, ensuring a secure and responsible AI environment.

Interested to see Atomicwork in action? Sign up for a free trial and let us show you how you can drive employee-led growth and business impact with our unified AI solution.

No items found.
Get a demo

You may also like...

Switch to Atomicwork: 8 key reasons to ditch your old IT support and service desk
From AI-generated answers to employee questions we’ve got a lot that you might be looking for in a modern IT support helpdesk.
The mission of Atomicwork
The story behind why we're building Atomicwork and how we're planning to empower employees every day with our products.
Atomicwork – on the journey so far, and the way ahead
Why we decided to start Atomicwork and work together for better technology experiences.