You’ll be surprised at how easy it is to create an effective AI assistant. With the release of large language models like GPT-3.5, building an AI assistant that can understand and respond to natural language queries has become much more accessible.

In this engineering blog, we'll explore how Avenue engineers built Copilot, an AI assistant that helps users create and manage monitoring tools and ask questions about our tool using GPT-3.5. I'll dive into the technical details of how we designed Copilot, discuss the challenges we faced, and share some tips and tricks that can help you create your own AI assistant.

Large Language Models (LLMs) are artificial intelligence models that are trained on massive amounts of natural language data, allowing them to generate human-like language and understand natural language queries. GPT-3.5 is an example of an LLM. These models have recently gained popularity due to their ability to perform a variety of language-related tasks, such as translation, summarization, and question-answering. In fact, this blog post was written by one of these models!

LLMs can be used to create AI assistants like Copilot, which can understand and respond to natural language queries. AI chatbots make it easier for users to interact with complex systems and save users time onboarding to your product. However, LLMs are not perfect and have some limitations. They can sometimes generate responses that are nonsensical, which is usually referred to as hallucination. As with any AI system, it is important to verify the information provided by LLMs, especially when dealing with business-critical tasks.

Another important limitation of current LLMs is that they can only handle a relatively small amount of tokens. For example, GPT-3.5 can only handle 4,096 tokens per request and response. Although it may seem like a lot, later on you will see that it actually imposes some significant constraints on our prompt.

 of a word that OpenAI’s models use to predict the next word. A helpful rule of thumb is that a token is about 4 characters, which is what I usually use to estimate the number of tokens my prompts have. Check out OpenAI’s blog post 

“What are tokens and how to count them”

Since we have such a small number of tokens the amount of contextual knowledge we can give our assistant is fairly limited. However there are some ways we can deal with this, which we will explore later in this blog.

For those who don’t know, a prompt is our primary way of interacting with an assistant and improving its performance. A prompt is a block of text that we provide an LLM to instruct it to perform a task. Here’s an example of a summarization prompt in OpenAI’s playground:

Prompts can be very simple. For relatively lightweight tasks, such as summarization, this may be all that you need. However, as your needs grow, the complexity of your prompt will also increase.

Lets look at an example of a basic AI assistant prompt for use with 

The following is a conversation with an AI assistant. The assistant is helpful, creative, friendly.

You’ll notice that this prompt starts with defining the role the model plays as well as some initial messages separated by “Human:” and “AI:”, which helps the model understand the flow of the conversation. This is a decent prompt for a very simple chat assistant that can answer basic questions. For example:

This prompt will give you a very general assistant that is similar to ChatGPT. For your use case you will definitely want something more specific.

To make it easier to construct your prompt, and to humanize the algorithm, I find it helpful to make analogies to human thought processes. My analogy starts with breaking the prompt down into a set of discrete parts that I call 

 is simply the name, personality, and role that the AI assistant takes on. The start of the example prompt given above is a good example of an identity that could be given to an assistant and may not need to be much more complicated.

 is institutional knowledge about your product inserted into the prompt. Knowledge 

 be added to a model with training or refining your own language model. But this requires a large training set to be done effectively and may not always be feasible, especially on your first implementation. As well, if your product changes and your docs need to be updated, then you have to build a new training dataset and retrain your model all over again. However, by inserting knowledge as part of your prompt you can dynamically add 

 knowledge, at run time, by utilizing classification. Effectively giving you an infinite 

 is an extension of the assistants identity and can help steer the conversation. In the prompt for Copilot, our directive includes more information such as “Copilot may be asked by the user to help come up with ideas for new monitors or collaborate on a monitor idea (’Ideation’)”. I find it useful to separate the directive from the identity so that it can be changed at run time to better fit a user message using classification. This can help your assistant be more flexible in the type of questions you wan’t to service, while still giving it some structure to work with.

 is an extension of the directive. It is a set of example conversations between a user and the assistant based on its current directive. It can help give some more structure to the conversation and allows you to insert specific strings that can be parsed out later to add more direct functionality to your assistant. For example, in Copilot’s imagination, we instruct Copilot to add the string “RECOMMENDATION:” proceeding every monitor recommendation it makes. This allows us to parse the title and description from a monitor recommendation and provide a link for the user to click on which then creates a monitor draft. Similar to knowledge, this can be accomplished with a custom model or a refined model if you have the data available.

 is simply the log of a single conversation between a user and the assistant. LLM’s are stateless, so continuity during a conversation is maintained by adding each user message and model response to a conversation list. It should be obvious that after a while this will become a problem due to the limit on tokens I mentioned earlier. Luckily there are some tricks for dealing with this.

Let’s imagine a SaaS company called “Pequod,” which enables users to build interactive maps and dashboards for GPS location data. Pequod wants to create an AI assistant named “Ahab” that will help users brainstorm ideas for an interactive map dashboard and for general documentation related questions. Let’s start with the identity of the assistant.

Next, the static portion of our institutional knowledge may look something like this. Right now, Pequod currently only has one piece of documentation, which we can attach to our knowledge section.

Since we want our assistant Ahab to offer ideas for interactive maps our directive may look something like this:

You should note that at this point the assistant has no context on the user or their organization, so we instruct it to gather this information before it makes any recommendations. When giving your assistant instructions, you should encourage it to ask clarifying questions to minimize the amount of unhelpful hallucinations it may otherwise return.

To complement our directive, we include an example conversation between a user and Ahab for the assistants imagination. Here is an example "Ideation" conversation between Ahab and a customer where "H" denotes the human participant, and "A" denotes the assistant, Ahab:

 with an initial user message which is simply “Hello, who are you?”

If we combine all of our sections into a single prompt, we can test it out in OpenAI’s playground and see how well we did!

The playground is a great way iterate quickly and improve your assistants capabilities. Try experimenting with changing your prompt and your parameters to see what works best for you and your organization.

The above use case was made for OpenAI’s text-davinci-003 API for simplicity. The GPT-3.5-turbo chat API (which Avenue uses for Copilot) is different and you may need to modify the prompt slightly. The main difference is that instead of accepting one large block of text as the prompt, the GPT-3.5-turbo API accepts a list of messages that can have roles associated with them. The three possible roles are “user”, “assistant”, and “system”. The prompt sections we’ve created above can be added as system messages proceeding the user and assistants conversation. See OpenAI’s documentation on 

 for more information on sending requests to GPT-3.5.

Strategies for dealing with the token limit

Keeping in mind the GPT-3.5 token limit of 4,096 tokens, our Pequod example’s prompt is about 1,000 tokens. This leaves us with over 3,000 tokens for the conversation between the user and the assistant. While this may seem like a large number, it is very easy to fill this space with documentation sections.

If we wanted to add more documentation into our knowledge section to give the assistant more context on the Pequod platform, we would very quickly run out of space. The way we dealt with this problem in our implementation of Copilot was by using OpenAI’s “Embeddings” API.

allows you to convert text into a numeric vector, which is a token representation of the text. With embeddings you can easily calculate a similarity score by taking the dot product of two vectors.

We store vector representations of short documentation snippets (<400 tokens) and at run time we calculate similarity scores for the latest user message and the entire conversation for all of our documentation snippets.

We then pick the two documentation sections that have the highest similarity scores and dynamically insert them into the knowledge section of the prompt. This allows you to have a theoretically infinite knowledge base for your assistant, while keeping the total number of tokens small. And because we do this for each user message, our knowledge section changes throughout the lifetime of a conversation based on the users needs. See 

this tutorial in OpenAI’s documentation

 for more information on using their embeddings API.

If a user is having a long conversation with the assistant you may still run out of space however. Solving this is much simpler. You can use the same GPT-3.5 API to begin summarizing the oldest messages into a single system message. This is a lossy transformation and the assistant will start losing some context about older messages, but this should be sufficient for the majority of use cases and will stop the conversation from failing abruptly.

Keep in mind the limitations and potential pitfalls of current LLMs, and always verify the information provided by your assistant for business-critical tasks. With these considerations in mind, you can create an AI assistant that provides real value to your users and makes their lives easier.

Interested in learning how Avenue can help you brainstorm ways to solve ops issues, write sql, and answer questions using OpenAI's ChatGPT? Checkout our blog highlighting, Copilot, your 

all-in-one ops, analytics, and customer service teammate 

In this engineering blog, we'll explore how we built Copilot, an AI assistant that helps users create and manage monitoring tools using GPT-3.5. I'll dive into the technical details of how Avenue engineering designed Copilot, discuss the challenges we faced, and share some tips and tricks that can help you create your own AI assistant. So buckle up and get ready to embark on a journey into the world of NLP and AI assistants!

How Avenue Engineering built an Ops Copilot with OpenAI

Avenue helps operations teams do their best work. That’s why we created Avenue copilot, your 

all-in-one ops, analytics, and customer service teammate!

 Powered by OpenAI’s ChatGPT, copilot can help you brainstorm ways to solve ops issues, write sql, and answer questions about Avenue.

You can start interacting with copilot on our app by visiting 

 or by clicking the “copilot” link in the apps sidebar. No set up required!

We are releasing our AI assistant copilot, which can answer a variety of questions you may have and collaborate with you on creating monitors.

Copilot has no context on your database schema or the specific SQL server you are using. To get effective support for this, make sure to provide Copilot with detailed instructions and consider providing specific table and column names that you know you’ll need for the query. It may also be helpful to explain any foreign key relationships that may be relevant. If you don’t provide this information, copilot will attempt to provide you with a decent starting point for your query, but you will have to fill in your table and column names.

Are conversations with copilot moderated?

Yes. Conversations with copilot are sent to OpenAI’s API and thus are subject 

. Your messages to copilot are not stored by OpenAI but they will be preprocessed for content that does not align with their guidelines. In general, think of copilot as your coworker, don’t say anything to copilot you wouldn’t say to a coworker.

What happens to my conversation data when interacting with copilot?

 sent to the OpenAI api for processing, but they 

 stored by OpenAI or used in any future training sets by OpenAI. See their 

 save encrypted versions of your messages that we may use for refining copilot in the future.

 with titles and descriptions that the assistant suggests to you. copilot can also help you create SQL queries for your monitors, but you’ll need to copy and paste them for now. Look out for future updates to copilot's monitor editing capabilities.

Not yet. But look out for more updates to copilot, including new ways of interacting with the assistant!

copilot is very capable and can help you with many tasks. However, current LLMs like GPT-3.5 are not perfect and are subject to 

. This means that if you ask copilot a question that it cannot answer based on the context it has, it will try and make up an answer. We’ve tried to encourage the model to communicate when it is unable to answer a particular question, but it can still happen. While copilot is certainly a useful tool, for business critical tasks it is always a good idea to verify the information copilott is giving you.

Interested in making an AI assistant for your platform? See our engineering blog 

Avenue helps operations teams do their best work. That’s why we’ve created Avenue copilot, your all-in-one ops, analytics, and customer service teammate! Avenue copilot, powered by OpenAI’s ChatGPT, is your newest ops teammate. It can help you brainstorm ways to solve ops issues, write sql, and answer questions about Avenue.

Avenue launches virtual teammate powered by ChatGPT

Avenue users can now build one monitor that is triggered by multiple data sources.

Data enrichment is a process that involves combining first-party data collected from internal sources (such as subscriber forms) with data collected from other internal sources or third-party external sources.

This feature is useful when you need to join data from more than one source. For example, if you had customer data in one data warehouse (like Snowflake or Redshift) and you wanted to join to outbound email data (from Front or 

). This feature is helpful for companies whose data sits across more than 1 database.

1.  A business collects data from a wifi device that is stored in their data warehouse, while their onboarding and customer data are saved in a separate database. By utilizing the multi-data source feature, they can seamlessly merge data from both sources and gain deeper insights.

2.  Another organization has real-time data stored in their postgres microservices, but broader store information is saved in Snowflake. By integrating data from both sources, they can gain a more complete view of their business operations and make informed decisions.

3.  A ride-sharing company stores device data in Presto and driver information in Hive. By leveraging the multi-data source feature, they can merge data from both sources to obtain a more comprehensive picture of their business.

To make it easier to ask for specific information, you can use the 

 Handlebars helps you organize your data so you can find what you're looking for more easily. When you use handlebars, the data is put into a context variable which contains all of the rows of information from your other query. This variable organizes the information so it's easier for you to read and understand.

To access a value in the first row, you can use the Handlebars syntax in the SQL Query for the enriching database. The syntax would be {{ data.[0].column1 }}. The data variable contains all the rows of information from the other query and is organized using context variables. By using this syntax, you can easily retrieve the value of column1 in the first row of data from the enriching database.

Let’s assume that the original query returned the following results:

If you wish to select rows in the enrichment query based on matching

 values, you can achieve this with the following approach:

This query will help you retrieve all rows of information from a table in a database called "

" that match a specific criteria. The criteria is based on whether the id value in each row matches any of the originalId values in a list of data.

Which handlebars would then compile into:

Avenue Launches Data Enrichment: Use Multiple Data Sources for Alerting

 is a digital pharmacy serving patients in Mexico since 2019. As a pharmacy and fulfillment center, the Vitau team is responsible for everything that happens to ensure that patients receive their medication on time. We spoke to Alejandro Lozano, Vitau’s COO, about how they use Avenue to power their operations command center.

As COO of an e-commerce business, Lozano needed to make the seams between the online and offline world less messy. Vitau had built their own Order Management System (ERP) so they had full traceability of their data, but lacked a pulse on business performance. They were stuck in reactive-mode. Before Avenue, Vitau was managing operations by blocking time on their calendars to constantly scan dashboards in Metabase and living in constant fear of the next ops fire. 

Avenue as the all-in-one command center for ops

Lozano needed to know when issues arose and who was accountable for fixing them. Let’s run through all the 3 types of monitors Vitau’s COO set up to have peace-of-mind and scale his operations team’s capacity 10X. 

These monitors trigger alerts to operators and escalate to key decision makers that something is not working as expected. They detect and surface problems in real-time, such as:

These monitors provide visibility across operations (fulfillment, payments, customer service, competition, pricing) to give Lozano assurance that the business is operating as expected. Vitau set up these monitors to signal when:

Proactive Push: The "Above-and-Beyond" Monitor

you should do these things to maintain good business health

These monitors trigger proactive workstreams for the team and allows managers to monitor how long tasks take operators to complete. When a monitor fires, a member of Vitau’s ops team is assigned a task of :

Vitau depends on Avenue to continually monitor for potential problems, flag and escalate issues to the proper people, and maintain a system of record for how they handle and manage these ops issues. “With Avenue as our supervisor we don’t have to seek out problems anymore,” says Lozano. Vitau has saved 10+ hours of operator time a week by eliminating “check for issues” meetings. 

Avenue has helped Vitau improve their operational performance throughout the company - increasing customer satisfaction, improving on-time deliveries, and reducing fraud

Interested in what Avenue can do for your company? 

How a digital pharmacy uses Avenue to power their operations command center

We want to help non-technical users understand the SQL logic powering their monitors. With the “Explain Monitor” button, users can now see in plain text what their alert does. Think of this button as “Google Translate” but instead of English-to-Spanish, it’s SQL-to-English. These SQL descriptions are AI-generated using ChatGPT so don’t expect us to win any Pulitzer Prizes with them. But, we hope they help you translate your company’s SQL logic into plain text so you aren’t waiting around for engineering or analytics teammates to unblock you!

Avenue is excited to announce the "Explain Monitor" button. Now you can translate your SQL queries into plain text using ChatGPT!

Avenue introduces “Explain Query” button to translate SQL

Avenue is excited to announce we now support subscriber notification filters for our Slack integration as well as our ticketing systems integrations (Zendesk, Asana, etc.). Users will now be able to specify exactly which signal updates they want to receive from Avenue when creating a monitor. Additionally, users are able to add individual subscribers to a signal after it has been triggered from a monitor run, either via the signals page or as a monitor action.

When creating a new monitor, users will notice the Slack and Ticketing integrations (Zendesk, Asana, Linear, Trello, Jira) will have an expandable menu that supports both notification body formatting and notification filters. Default options for each supported subscriber are already selected, but users are able to adjust the settings as they see fit.

Depending on the type of integration being added as a subscriber, the type of notification users receive will be different. For example, Slack status update and assignee update notifications are delivered in the signal’s slack thread. However, for ticketing integrations such as Asana, status and assignee updates are reflected via the corresponding fields in those ticketing systems, such as the Asana ticket status and ticket assignee.

Once a monitor signal has been triggered, any integration subscribers will now appear the signals page in the bottom right corner. The subscribers can be clicked to bring the user to the external system, and the active notification filters can be seen.

If a user wishes to update the active filters for a subscriber, simply hover over the subscriber and click the edit button, and select the new set of filters to apply!

Adding additional subscribers to a Signal

Prior to today’s launch, Avenue users were only able to specify which subscribers to attach to a signal in the monitor edit form. The subscribers would be auto-added to new signals only, anytime a signal was triggered by a monitor run. Now, users are able to add one-off subscribers to signals 

 a signal has been triggered via the new 

 button, directly from the signals page. For now, only subscribers to Slack and Ticketing systems are supported, with support for more integrations (such as WhatsApp, SMS, Email) rolling out in the future.

In addition to adding a new subscriber from the signals page, users are now also able to add subscribers as a monitor action, triggered via slack or via a an event. Simply head over to the Monitor Action creation page, and setup the subscriber to be added when the trigger conditions are met.

Monitor subscribers now support notification filters, and users are able to configure these filters during monitor setup or after subscribers have been attached to a signal after monitor escalation. Additionally, users are able to add individual subscribers to signals directly from the signals page, or through a monitor action via an event or slack action trigger.

Will I be able to remove a subscriber from a signal?

Avenue currently does not support the ability to remove a subscriber from a signal.

When adding an individual subscriber to a signal, will all subsequent signals also receive that subscriber?

No. Whenever a signal is created from a monitor run, only the subscribers defined on the monitor create form will be attached to the signal. Any subscribers added to a signal after the monitor run are specific to that signal only.

When adding an individual subscriber via the signal page or via a monitor action, why are SMS, WhatsApp, Email, and Webhooks not supported?

At this time, we are unable to support all subscribers configurable via the monitor creation form as individual subscribers. In the near future, we hope to change this limitation!

Avenue is excited to announce we now support subscriber notification filters for our Slack integration as well as our ticketing systems integrations (Zendesk, Asana, etc.). 

Avenue expands features for Signal Subscribers

CommandBar is an intelligent layer that sits on top of any app and adds widgets that guide and assist end users to help them wield software more powerfully. These widgets include a magic search bar for commands, and nudges to suggest next steps and redirect users. Developers copy a code snippet to install CommandBar and create and edit experiences using a low-code editor. An analytics dashboard on the back end shows the most popular user behaviors, which developers can use it to improve their product and configure new experiences in CommandBar.

Lucas Howard joined CommandBar as Employee #1. Being the first hire at a hyper growth startup means that Lucas has worked on multiple areas of the business. He’s worked on operations and currently focuses on product and analytics. Lucas works with several stakeholders in the organization from executive leadership to the operations, product, and engineering teams.

Avenue is a key pillar of CommandBar’s operations

At CommandBar, Avenue touches on all aspects of the business. From understanding user behaviors to having visibility when something needs to be attended to resolving those issues quickly, the team cannot imagine a world without Avenue.

Operations teams fill the gaps where the product falls short. Using 

Avenue has helped the ops team at CommandBar catch scenarios that are not supposed to be happening

 or configuration issues on the client side.

“Unless you check dashboards manually you won’t see the data”

Very early on, the CommandBar team realized they had a ton of interesting product data they wanted to do more with. The founding team is very technical, so there has always been a strong focus on data and analytics. After doing a series of sprints to build data infrastructure, the team set up tools like DBT, High Touch, Stitch, Redash, and Retool.

“unless you check those dashboards manually you won’t see the data”

. This process was reactive, not proactive.

Because “Slack is the heartbeat of the company”, the team at CommandBar found it extremely valuable that these metrics were coming into Slack throughout the day. Now, user data comes into their Slack in real time which 

 about how users are using the product and any problems they may be experiencing.

Proactive Customer Support: “Avenue allows us to be proactive by helping us uncover users who may be at risk for churn”

More than just general usage insights, there are 

 that can be the difference between a customer having a 

successful trial or them concluding the app is buggy and churning

For example, sometimes when a company is inviting new teammates, the new user may sign up individually and create a new “organization”. This leads to duplicate organizations and causes the new user to experience confusion. 

CommandBar solved this problem using Avenue

, Lucas created a query to find duplicates and set up alerts to let the customer support and sales team know when this happens. Now the 

CommandBar team reaches out proactively to resolve this problem

. Less confusion means higher likelihood of conversion and retention.

Avenue touches on some of CommandBar’s most important KPIs

Operations teams have been the group that works on the edge of a business. They pick up everything that slips through the cracks, and they tend to do so with a patch-worked and insufficient toolkit. Instead of focusing on preventing fires, they have to figure out where the fires are coming from while they fight them.

It’s a new day for operations teams. Avenue provides all the alerts, monitoring, analysis, and escalations you need, in one place.

How CommandBar uses Avenue to build product intuition and optimize engineering resources

Building product intuition and optimizing engineering resources: A Case Study with CommandBar

2022 was a huge win for Avenue. We saw our team grow, customers scale, and our product improve.

Looking forward to 2023, we’re excited to continue on this trajectory and find ways to exponentially add value to operations teams.

Below are some of our most notable milestones and features from this year! Thank you to all who have played a part, big or small, in our journey so far.

2022 was a huge win for Avenue. We saw our team grow, customers scale, and our product improve. Here are some of our most notable milestones and features from this year!

Avenue 2022 Roundup

As the operations team at your organization, you are constantly building systems, which often boils down to coordinating people across the various tools they live in.

That's why we are excited to announce that we have launched a new suite of integrations with popular ticketing tools like Linear, Jira, Asana, Trello, Zendesk, and more.

With these new integrations, you will be able to easily create a playbook in Avenue that can

Avenue enables teams managing each playbook to focus on solving each step the instant they become known. In some instances, an Avenue signal may surface the need to create some followup work in an external issue tracking system. Previously, users would manually transcribing Avenue signal details to these external systems. After we launched our Zendesk integration, Zendesk quickly became one of our top used integrations and received positive reception. The most popular request we received after the Zendesk launch was support for more issue tracking integrations similar to Zendesk. Today, we are excited to announce that Avenue users can now automatically generate tickets in Linear, Asana, Jira, or Trello, directly from Avenue. Easily execute your team’s full response playbook by including this in your workflows!

To get started, head over to the settings page and scroll to the bottom.

Next, click the new integration card and supply a name for the integration

We use Merge API to handle authentication and communication with Asana, Jira, Linear, and Trello. Our Slack, Sendgrid, 

, and Zendesk integrations are not powered by Merge. Follow the steps in the Merge Link UI, and when you’re finished, you should see the integration card appear in the integrations section.

Finally, navigate over to the monitors page, and either create a new monitor or update an existing one. Scroll down to the bottom, and your new integrations should be ready to configure as notification destinations

Integration support for Asana, Linear, Jira, and Trello ticketing systems.

What features are supported by the new Integrations?

Similar to our Zendesk integration, users will be able to configure monitors to create tickets in these external systems when a signal is triggered by a monitor run. Users can also expect Avenue comments, assignee updates, and status updates to propagate to the external ticket. Commenting on a ticket in the external system will also sync back the comments to Avenue as well, and by extension, any dependent subscribers for that signal.

If I comment on a signal or update the signal state from Slack, will that propagate to the external tickets as well?

What features are not yet supported by the new Integrations?

File uploads are not yet supported, but will be coming soon. Additionally, updating the status or assignee of an external ticket will not propagate back to Avenue and related subscribers. At the moment, only comments are kept in sync.

Why are my comments not showing up in Avenue when I comment on an external ticket

For our Asana, Linear, Jira, and Trello integrations, comments are synced back to Avenue once every

. In the future, we hope to bring this number down to a maximum of

Will I be able to disable what updates are synced from Avenue to the ticketing destinations?

Not immediately at launch, but within the week we will release the ability for users to determine what changes should be propagated from Avenue to external destinations. Stay tuned for more updates!

Avenue now supports creating and syncing work across all your ticketing tools

Avenue Launches Support for Linear, Asana, Jira, Trello, and More

Avenue now offers Audit Logs for Monitors, allowing users to see a timeline history of all the creation, deletion and interim changes performed by anyone at the organization to a given monitor.

Monitors change over time. They're initially created with one intention, thresholds are then modified, queries updated, they get paused/unpaused, and more. 

Previously, there was no easy way to track your team's changes to a given monitor. This meant changes would go unnoticed or unexplained.

Now, we have two places you can observe a monitor’s history. The individual monitor audit log, located on the monitor details page, and the global audit log, located under the settings tab of the navigation bar.

Both areas surface monitor events, the distinction being that the Monitor-level Audit Log only shows the history for that single monitor, and the Organization-level Audit Log allows you to filter through and view changes across all monitors belonging to your organization.

Both are also now live! Simply update your monitors as you normally would and the changes will be reflected in the locations mentioned above.

Two monitor logs detailing changes and updates related to your organization’s monitors. One global event log, located in the navigation bar, and one local one for each monitor.

No. The audit log will only contain events that have occurred since this feature was launched.

At this point, logged updates include: monitor creations, monitor deletions, monitor status changes, updates to certain fields within the monitor itself (like query, title, description changes), monitor group changes, and monitor auto assignee changes.

Will additional event types eventually be added?

Over time, we do intend to expand our events types further, namely including more of the changes that can occur on the monitor edit form. However, this will likely be the full scope of even types for the time being.

Events should show up in both monitor audit logs almost immediately after they occur.

Can I get these events sent to me as alerts?

Yes! If you are an admin of your organization, you can manage an audit log subscription via a button on the upper right hand side of the global organization audit log. Currently, you can only get them sent as Slack alerts, but we do intend to expand that to emails, texts, etc over time.

Avenue now offers Audit Logs for Monitors, allowing users to see a timelined history of all the creation, deletion and interim changes to these entities.

Avenue Launches Audit Logs

Operations teams almost entirely find work to be done in one of two ways:

Avenue has historically solved for the first use case - giving Operations teams their time back so they can spend more time solving vs. searching. Today, with the rollout of Forms, we’re excited to start supporting Ops teams beyond proactively monitoring on their behalf and support them in manually-triggered Playbooks and Workflows.

Avenue Forms introduces a new paradigm to the Avenue Alerting ecosystem. Our customers currently use Avenue to passively monitor their data warehouse. However, we've found that Operations data does not always start and stop with the data warehouse. Often operations data lives outside of your virtual ecosystem. That's why we've created Avenue Forms to give you a way to manage your internal operations on the fly!

Avenue Forms are flexible and can fit a wide variety of use cases. Internally, we use forms to surface feature requests and to help manage our customer support pipeline. Some of our forms-beta users have been adopting Avenue Forms to manage Delivery-Support (driver calls with a failed dropoff reason) or when their in-store staff have an issue to flag.

When can I get started with Avenue Forms?

 Avenue Forms is available now in your workspace. Go to the Forms sidebar item to get started.

You can view and submit your Avenue form by triggering the Avenue Slack bot with a 

 command or a message shortcut then select a webhook monitor with an attached form from the dropdown.

Can I send a link to an Avenue Form to my customers/non Avenue users?

Forms are currently only available to users logged into your Avenue workspace. We do, however, have plans to enable forms to be sent to non authenticated users in the near future.

How does this feature affect my current webhook monitors?

 All your current monitors will be unaffected by this change.

Today, Avenue Launches Forms - allowing Operations Teams to easily create a form that can be used to kickoff a Playbook. Now, Avenue can be used to not only to passively monitor all possible problems, but can be used by Ops teams to kickoff proactive workflows too.

Latest posts