Any engineer that has had to debug an issue in production will know that logs are a crucial component of the dev-ops toolkit. They’re almost certainly the first place you look when you’re notified that something in your app is broken, so the quality of your logs is highly correlated to the time it takes to diagnose, and ultimately fix, any issue that you run into. The only thing that can make a 2 AM on-call alert worse is the subsequent realization that you’re not going to be able to root cause the problem because it was not adequately logged. In an effort to help you avoid that situation, let’s go through the core traits of a successful logging solution below.

In most cases, a log should be as close as possible to the source of the error. This allows you to highlight the problem with the highest level of fidelity. As you bubble the error up through your call stack, you become more likely to conflate multiple sources of error into one generic log message. Ultimately it’s the difference between knowing that your API failed, and knowing the specific external dependency that caused said failure.

A log message should, at the very least, always contain the contents of the error that it is logging. In the example below, the log message on the bottom is exponentially more useful than the log message on top because it will provide the reader with substantially more information about the root cause of the failure:

Knowing the answer to these types of questions is fundamental to knowing how to fix the issue.

// Bad
logger.error("API request failed")

// Good
logger.error("API request failed", {
	errorStatus: err.status,
	errorMessage: err.message,
	errorStack: err.stack
})

Today your logs are most likely aggregated into some persistent storage like DataDog, Splunk, or AWS CloudWatch. These log stores are fantastic tools for combing through the mountains of logs that your servers emit, but they’re far more useful if your logs are well-structured. In production you should emit your logs in a format like JSON so that your aggregator can properly index your logs. Likewise, rather than emitting a log as a monolithic string, emit it as an object with a key-value pair for each relevant piece of information.

// Bad
{
	"content": "API request failed with code 429: rate-limited"
}

// Good
{
	"content": "API request failed",
	"errorCode": "429",
	"errorMessage": "rate-limited"
}


This will allow you to explicitly search for each field, instead of always doing a fuzzy text-search on the content.

Lastly, a log message should be contextually rich. The definition of context will depend on your application, but it should ultimately serve to semantically identify a given log message from other log messages that may have been emitted at the same time or place in your app. In a nightmare scenario where you’re suddenly flooded with a series of errors, the first thing you’ll see is a chronologically-ordered glut of interleaved logs coming from a number of different places in your app.

It's incredibly important to have the capability to filter down these logs to a 

 of a given request, workflow, or other process in your app so that you can make sense of the noise.

Ultimately, identifying and agreeing on the aspects of a good log message is only useful if those qualities are easy for the engineers to implement. Location, content, and structure can mostly be enforced through documented best practices and code reviews. However, making sure that each log message has adequate context is a non-trivial problem. Let’s use Avenue’s distributed workflow engine as an example.

Avenue’s workflow engine is a multi-tenant application that runs thousands of cron jobs every hour to evaluate the conditions of various Avenue monitors, trigger escalations, and more. For a given job, let’s define our base context as a 

. It’s impractical and tedious to independently resolve these values every time we want to emit a log message. A natural alternative is to pass around a context object with these values throughout our entire call stack. This is the approach 

// Foo takes a Context object as an argument
function Foo({ctx, arg1, arg2}: {ctx?: Context, arg1: any, arg2: any}) {	
	// Foo uses the context to instantiate a logger and emits a log
	logger.fromContext(ctx).info("This log includes context!")

  /*
		The resultant log looks like this:
		{
		 	"content": "This log includes context!",
			"jobName": "Foo",
		 	"monitorId": "monitor-1",
		 	"signalId": "signal-1",
		 	"organizationId": "org-1"
		}
	*/

	// Foo takes the Context argument it received, and passes it on to the
	// the next function it calls
	Bar({ctx, arg1})
}

// Bar also takes a Context object as an argument
function Bar({ctx, arg1}: {ctx?: Context, arg1: any}) {}	


The results are traceable and consistent logs, but the implementation still feels a little clumsy because you are required to pass a context object throughout your entire call stack. Luckily 

 provides us with a way to achieve the same end product without any boilerplate!

OpenTelemetry is a "collection of tools, APIs, and SDKs [used to] instrument, generate, collect, and export telemetry data (metrics, logs, and traces) to help you analyze your software’s performance and behavior.” It’s a 

, and an altogether fantastic collection of software. We can use OpenTelemetry’s NodeJS Context Manager to register and propagate our context object down our call-stack without explicitly passing the object to every function!

The magic behind this module is Node’s concept of 

, which allows asynchronous resources (such as Promises and functions with callbacks) to inherit contextual information from their “parent” resource without explicitly passing around a map to store this metadata. If you’ve used 

 in other programming languages, this concept should feel familiar!

In practice, the developer experience is as follows:

Thus we can instantiate context once at the top of our call stack, provide it to the OpenTelemetry context manager, and leverage it whenever we need to emit a contextually rich log message. Adding context to upstream functions is seamless, and similarly, updating existing log lines to leverage this context took minimal refactoring. If you’ve made it this far, hopefully this pattern can help you bring a bit more structure and consistency to your logs! If you’re interested in learning more about the various applications of async local storage, async hooks, and persistent execution context, this 2020 talk by Shanee Radzewsky is an awesome resource.

Async Hooks: A Journey To a Realm With Persistent Execution Context | Node.TLV 20 | English

Likewise if you’re passionate about observability, building resilient software, and are looking for something new - Avenue is hiring across engineering, design, growth, and more!! Check out our 

How Avenue Makes Effective Logging Effortless with OpenTelemetry

New to Avenue: now you can add tags to any Monitor. You can use tags to categorize Signals based on common traits and themes.

Companies scaling their Operations with Avenue end up with a long Signals page that covers several work-streams. As a result, it can be difficult to keep track of common themes. Now, you can tame the chaos with tags. Simply add a tag in the Monitor or Signal Details page and watch them appear in the list views.

Tags are live today! You can create and add them directly in a monitor or signal details page, and edit them in user settings.

Tags to link common traits in monitors and signals.

You can create a tag either directly from the monitor or signal details page by inputting your desired title and clicking the “+” button or in the tag section of the user settings page.

You can only edit tags in the tag section of the user settings page. Tags cannot be edited or deleted in the monitor or signal details page.

You can only delete tags in the tag section of the user settings page. Tags cannot be edited or deleted in the monitor or signal details page.

Do signals and monitors have different tags?

No, tags are consistent throughout both signals and monitors. When you create a signal, it inherits all of the tags currently on the monitor it was created from.

For one level up, try using entity tagging!

Avenue Launches Monitor Tags

Avenue users can now automatically generate 

Teams managing incident response should be able to focus on solving problems. With Avenue, users can automate the human coordination and documentation of that problem solving, including auto-creating and auto-populating Zendesk tickets. Easily execute your team’s full response playbook by including this in your workflows.

Skip the manual work and focus on handling the emergency - we’ll handle the rest.

To use this custom mapping tool, simply navigate to the Monitors page > Create Monitor, and select “Zendesk” as the Notification Type. From there you can map the Monitor results to Zendesk fields.

Does Avenue intend to roll out support for more third party APIs?

Yes! Stay tuned for a few more and check your Avenue workspace’s settings for a sneak peak.

What kind of content can be placed in the mappings?

Users can inject data from their monitor runs as well as hard coded content.

Avenue Launches Custom Actions in Zendesk

 works with trade and technical schools across the country to provide affordable payment solutions for students. They build, manage, and fund custom tuition solutions for high quality trade and technical schools that prioritize student outcomes. Through their specialized software and human support, they facilitate custom Single Payments, Income Share Agreements (ISAs), Payment Plans and Hybrid structures. Mia Share gives more students the ability to pursue affordable, valuable education while helping partner schools improve enrollment, retention, and outcomes.

Understanding Income Share Agreements (ISAs)

The team at Mia Share is fiercely devoted to making education financing more fair by providing alternatives to student loans. Under an income share agreement (ISA), the student receives their tuition in exchange for a fixed percentage of their income. There are constraints around what the student owes each month. If they don’t earn above a set income threshold each month on an annualized basis, they won’t owe anything; only when they surpass the income floor do they owe a percentage of earnings back per month. On the other end of outcomes, the ISA has a cap for money owed, and if the contract ends before the tuition is paid back in full, the agreement is still terminated without any remaining balance or accumulating interest.

The Operations Team at Mia Share had a long list of alerts they wanted to receive, but it meant submitting requests to Engineering for every change, even if it was only to update the Slack channel where they wanted to re-route the alert. That’s when they implemented Avenue.

“The queries are easy to replicate, and I make quick changes to create a new notification,” explains Molly Roberson, Head of Operations at Mia Share. “On the day-to-day, having Avenue gives me tools to keep an overview and understand the system more. I can’t constantly bombard our engineers with new notifications or small changes. I also get to set up notifications when I’m onboarding new team members to help familiarize them with different processes and aspects of our company.”

Mia Share’s Chief Technology Officer, Derrick Ontiveros, was hesitant at first to implement a tool like Avenue. “I was concerned that it wouldn’t be accessible for non-developers and skeptical that it would save us time, but it really has,” he shares.

Before Avenue, Ontiveros would have to allocate resources to alert creation, increase the complexity of Mia Share’s internal systems, and continually manage the list of alerts for Roberson and her team. Now, they have notifications “up and running more easily with Avenue than anything else.”

With their focus on simple, smart, and scalable technologies, the Mia Share team is ready to grow. Internally, they’re focusing on growing the team and continuous product improvement. Strategically, their aim is to increase their product offerings for partner schools and expand to a wider range of trade and technical schools across the country this next year. Underlying both goals is their core mission: “Our goal is to continue making education more accessible to students and to continue giving them alternative options to student loans.”

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 do so with a patchworked 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.

Independence for Operations Teams: A Customer Case Study with Mia Share

What are you responsible for at Mia Share?

I’m responsible for managing several teams, including Student Success, Content and Design, and Operations, where we have Avenue helping us. Part of that operations function includes overseeing the process of underwriting, creating contracts, making changes in the system, and so on.

A way to summarize being in charge of Operations is making sure that things go well across the board.

What are your biggest goals and operational challenges today?

My biggest goal is getting everything streamlined. Everything we do should be very organized and easy to use so the company can hit their larger goals.

The biggest challenge as a startup is that you’re growing rapidly. Not only does it mean your customer base is expanding quickly, but so is your internal team. Being able to train and onboard new employees properly and make sure everybody is on the same page is a really exciting challenge that we’re facing. Implementing Avenue has given me the ability to quickly set up alerts that we need and not rely on our tech team in order to get that done, both for customer facing alerts and employee onboarding. On that end, I’ve been using Avenue here to set up individualized alerts for new employees to help them get acquainted with different processes.

Our goal is to continue making education more accessible to students and to give them alternative options to traditional student loans.

I’m most excited to do this in an innovative way. We don’t want the servicing of these payment products to be similar to servicing of the past. I’m really excited to find out how we can be successful in a more interesting and student-friendly way.

My advice would be to stay organized and think long term from the very beginning. Stay on top of your strategy and the overall goals you’re trying to hit, and don’t get lost in the day to day.

Q&A with Mia Share’s Director of Operations Molly Roberson

Fireside Chat with Mia Share's Director of Ops Molly Roberson

Avenue users can now take custom actions with 

Go beyond the alert. With Avenue you can automate the human coordination of how problems get solved in your organization, all the way through whatever action might resolve the issue (like triggering a 

Skip the manual work and focus on handling the emergency, we’ll handle the rest.

Use Avenue to easily execute your team’s full response playbook by setting up workflows. Avenue notifications can now be escalated directly to creating a 

 campaign, by auto-populating the desired fields and triggering the send.

To use this new customization, first visit your profile settings (

https://app.useavenue.com/account/settings

. Then, in any Monitor, simply select “

” as a notification endpoint, and fill out the form that appears before you.

What third party APIs are currently supported?

Are webhooks supported as trigger origins as well as SQL queries?

Avenue Launches Custom Actions in Customer.io

 is a grocery automation service that takes care of the shopping, delivery, and meal-planning for its customers. Your mission control for food, Jupiter makes it easy to do your weekly shopping with personalized recommendations and delivery to your door. Their autopilot technology learns from your preferences to help you build your cart and spend less time planning and shopping each week. 

I’m the Co-Founder & CEO of Jupiter. I’m involved in all of our teams, but my days are focused more specifically on Product.

Product distribution. It’s the most important thing to survive and build a company in today’s world if you’re an e-commerce, food distribution company. The first component of that is national expansion. The second is thinking about acquiring users more organically.

Another goal is when I think about Jupiter being a real network. All food creators should live within the e-commerce experience. If you see a video, you should be able to buy that meal. If we can embed Jupiter behind every piece of food content on the Internet, then we’ve really unlocked the next version of e-commerce. That’s the goal, and to get there quickly.

What tools or processes are you using today? 

We’re structuring the company in a way where our systems are set up to scale. We have a support system that we have an outsourced company managing, so as those needs grow, it scales with them. The other part of scaling is making sure engineers feel happy and confident scaling on top of this system. We talk a lot about tech debt to ensure productivity is high.

I’m really excited to see the results of some tests we’re running. What we’re doing is something the market hasn’t seen before.

Advice for other founders or people running Product/Ops teams?

Maintain a healthy amount of optimism. There is a lot of doubt, fear, and uncertainty in entrepreneurship. Optimism will pull you out of that uncertainty. Optimism is the oxygen for the team. If it’s gone, it’s hard to breathe. Build the house one brick at a time. As long as you keep learning and following the process, you’ll find the path.

To learn more about how Chad implemented Avenue to superpower Jupiter's product and operations teams, read 

Avenue sits down with Jupiter's Co-Founder & CEO Chad Munroe to discuss product distribution and advice for founders.

Fireside Chat with Jupiter's CEO & Co-Founder Chad Munroe

on Avenue: the easiest way to effectively manage your monitors and organize your dashboard meaningfully. You can now group your monitors based on similar attributes or metrics and take bulk actions on these groups with a click of a button.

To enter Edit Mode, click on the Edit button on your Dashboard. Here, you will be able to create new monitor groups.

Click on individual monitors to select, group and ungroup selected monitors, or bulk select “All” or “None” in the available dropdown. You can also bulk pause selected monitors. If all selected monitors are paused, you will be given the option to bulk resume. If, however, you select a monitor configured via Webhook, you will neither be able to pause nor resume the selected monitors.

Edit Mode is available immediately. Customize your Dashboard and take advantage of this feature today!

Edit Mode to group your monitors and take bulk actions like pausing/resuming.

Select which monitors you want to group, click on the “Group” button and then choose the desired group from the dropdown of existing groups.

To ungroup, simply select the desired monitors and click on “Ungroup”. You can ungroup monitors across multiple groups — this will just remove these monitors from their respective groups.

Click on the pencil icon next to the title to rename the group.

Click on the X icon on the left of the title to delete the group.

Does deleting a group bulk delete its monitors? 

No. Deleting the group will ungroup all monitors belonging to this group. It will not delete these monitors.

Why can't I see the bulk pause or bulk resume button? 

You have probably selected a monitor configured via Webhook! Such monitors are always active and cannot be paused or resumed and this is why you can’t see the bulk pause or bulk resume button.

Note: all non-Webhook monitors can be bulk paused. To bulk resume, make sure you’ve only selected paused monitors.

How can I unselect all selected monitors after I'm done editing them? 

We know how frustrating it can be when you select many monitors, accidentally click a button (for example, pause) and want to undo what you did, but all selected monitors have vanished! We won’t do that to you.

After taking an action, click on the dropdown next to the search bar and select “None” to unselect all monitors easily.

Avenue adds monitor grouping and bulk actions.

Avenue Adds Monitor Grouping and Bulk Actions

Today we're launching role-based access controls in Avenue. Now organizations can assign permissions according to the different user types, all within their Avenue workspace.

The end result is a more focused and customized experience for all Avenue users, and a more secure application for all Avenue customers.

The ability to assign access permissions for different types of Avenue users.

What are the different roles available in the Avenue app?

The rule of thumb for role assignment should be to give each member of your organization 

 permissions required for them to execute their role. That is if you anticipate that Bob will only respond to alerts created by others, make him a Member. Likewise, if you’d like Alice to be responsible for creating new alerts, make her a Creator!

How do I change my (or my teammates’) role(s)?

 page in the Avenue app and click on the pencil icon to open the editing pane for a given organization member:

This will open a pop-up where you can edit the selected user’s role:

After confirming your selection, the updated user’s role will be reflected in the Teams page and the next time they log into the Avenue app.

Now organizations can assign permissions according to the different user types, all within their Avenue workspace. 

Avenue Adds Role Based Access

 is building the operating system for grocery. A key part of our society yet extraordinarily underserved when it comes to technology, grocery stores have been making do with a patchwork of systems and band-aid solutions. The vision behind Vori is to be the trusted data layer within the global food supply chain. By streamilining orders and communication with vendors, Vori aims to democratize access to food supply and grocery data.

I’m the Head of Operations, which means I’m responsible for a wide range of functions that allow Vori to run and scale. That includes onboarding and launching new stores and vendors, managing customer success, and ensuring data integrity across all partners.

We are continuously iterating on making our processes more efficient and maintaining data integrity. Data integrity is what powers our platform.

I owe my operations acumen and skillset to Uber. My four years on the Freight team required me to be SQL proficient, highly analytical, and able to make my way around big, hacky spreadsheets.

My biggest goal — and my biggest challenge — is maintaining a high bar for reliability and service as we scale. People depend on us to get food to their communities, so failure is not an option. It is also not acceptable to have anything less than the best customer experience.

What is the primary focus of the broader company? 

Democratizing access to the local, national, and global food supply. That means being able to give retailers all the information they need to make the best decisions for themselves. That comes with a lot of responsibility.

Our technology is uniting the full life cycle of ordering inventory. We’re knitting together each of the various steps so that retailers, and suppliers, can better manage inventory and understand their stores. The primary focus of Vori is to create an inventory management system that is deserving of the amazing customers that we serve.

There’s something deeply emotional about the grocery store experience. A lot of people have fond memories of wandering down aisles with their parents or guardians. These retailers are often cornerstones of their communities, yet they sometimes lack even basic data and tooling.

A lot of times what people love about their grocery store are the items you can’t get anywhere else – the interesting SKUs. We’ve focused a lot on that, as the amount of overhead that it takes to support a range of suppliers is very challenging. Every one of them face regular pricing and catalogue updates. When it comes to replenishing shelves - and we have seen firsthand during COVID what happens when that doesn’t go right - our retailers don’t have all the information they need to make the best decisions for themselves.

I look forward to reinventing the experience of working at and owning a grocery store.

The bar is so low for customer experience when it comes to the grocery industry (and supply chain as a whole). Given that this space is so underserved, people often scoff at customer support functions because they wonder, 

“Are you really going to be there when we need you?”

Don’t be afraid to do non-scalable things that delight your customers, especially early on. To a certain extent, you can’t scale this, but when you can, you should.

The secret to Vori’s success has been hiring genuinely curious and radically diverse people. It’s incredible what you’re able to get done when you have the right group doing it.

To read more about how Maika implemented Avenue to superpower Vori's operations teams, click 

Avenue sits down with Vori's Head of Ops, Maika Koehl, to discuss scaling operations and customer support.

Fireside Chat with Vori's Head of Ops Maika Koehl

is building the B2B operating system for the grocery industry. A crucial part of our society, yet extraordinarily underserved when it comes to technology, grocery stores have been making do with a patchwork of systems, bandaid solutions, and analog methods. The vision behind Vori is to be a trusted data layer within the global supply chain and the standard for multi-player trade, communication, and prediction within society's foundational B2B industries. By streamlining orders and communication between grocery retailers and their suppliers, Vori is walking out its mission to unite the supply chain to ensure communities have access to the goods they need and love in order to live, work, learn, and play – starting with food.

As Vori quickly expands from a nascent startup to serving thousands of customers, the Operations Team’s scope of responsibility includes a wide range of functions that allow the company to run and scale:

Maika Koehl, Vori’s Head of Operations, recognized early on that her team needed to consolidate data and standardize knowledge.

“Vori has been live for 24 months, and we’re well beyond our first set of beta customers,” Koehl explained. “We’re a scaled team and we’re growing very quickly. As a result, we’re pulled in a dozen different directions. When I joined a year ago, I knew we needed someone — or something, like Avenue — to tell us what we should care about so we can focus our attention on what’s most important, when it’s most important.”

Growing Data Piles with a Growing Company

Similar to many operations teams, Koehl and her team were carrying out manual checks, building out massive Excel spreadsheets, and running to catch anything falling between the cracks.

“We would review the data in aggregate, but when you do that it can mask what’s happening, and you’re getting the information at a time when it’s not necessarily most impactful to reach out.” Koehl’s team was stretching their bandwidth, from manually checking orders placed on Vori’s platform to physically dropping into their customers’ stores and warehouses.

Koehl set up a collaborative process to come up with the use cases for Avenue. With a list of pain points and manual checks, her team came up with a shorter list of main categories for the first use cases. Koehl was then able to quickly deploy Avenue alerts for her team: “What’s nice about Avenue is that I’m able to create my own signals, or pass them along to our intern to help us with that.”

A guiding principle for Koehl and her team is to maintain the highest levels of customer support for the ultimate customer experience. They set up Avenue as the omniscient, omnipresent observer to trigger alerts to the team for customers that needed attention. One of the monitors Vori set up is to alert when a customer is at risk of churning or goes a certain amount of time without engagement, both at the store and user levels. Avenue also sends pings when a customer signs up or places their first order.

Before Avenue, the Vori Ops Team was scrupulously checking several Slack channels that constantly received pings. Now, those alerts have been cut by over 50% due to Avenue’s prioritization mechanisms. And those large spreadsheets? Koehl confirmed that all alerts have moved to an Avenue and Slack combination: 

Vori is growing quickly. With new product rollouts and pilot programs, there will be several launches at a time. “We're scaling nationally this year,” Koehl shared. “Being able to have not only Avenue’s granular pings that indicate success or failure, but also having the birds eye level view with the Avenue Dashboard is going to be a game changer.”

How Vori Scaled Operations with Avenue. Operational Analytics, Operations, Data, Growth, Dashboards, Spreadsheets.

How Vori Scaled Operations with Avenue

 allows investors to invest deal-by-deal in world-changing startups. From Rolling Funds to Syndicates, Roll-Up Vehicles and more, AngelList has made it simple for Fund Managers to Set Up, Raise, and Deploy.

Fund Managers can streamline their fund with AngelList's best-in-class software and tools. In the meantime, AngelList Ops Teams (Compliance, Data, Etc.) are handling all the legal formation, capital calls, tax documents, and all the other back-office service your fund needs, no matter its size.

How do you organize Operations at AngelList today?

There are many ways to staff an operations team. At AngelList, we break down by stage of deal: there's a team for setting up an entity to close, executing the work itself (LPs, GPs, documents), and a team for after a deal closes (post-closed actions whether it's a new round of financing, exit, or anything in between).

What are you responsible for at AngelList Venture? How has it evolved?

I started off at AngelList doing financial operations. This involved managing anything and everything that happens with companies, such as the exits and distributions, and the processes that come with that.

I moved on to work on data-related items because I recognized a gap at the time. As is common in the early stages of a company, we didn’t have a data team. I started a data request board and was pulling together people’s data requests. It ballooned into something much bigger, and it became clear that we really needed to use the data we were pulling and get it back into the system. 

With this realization, I worked with one of our data engineers to build a data stack, rolled out Metabase, and finally had a flurry of dashboards that informed us of every single product within the company and every single user flow. We got to the point where we had all this data, and I thought to myself: “What’s next? How can we can we use this to actually improve the efficiency of operation workflows?” And that is where BizOps comes in.

What was your focus when you launched BizOps?

I began leading the DRI’s trying to understand what 

 were they solving them. I worked cross-functionally with Engineering and Product to understand if they can build the solutions or implement something lightweight. In that respect, we realized we had to go the lightweight route, which is when we implemented Avenue.

Your role evolved from the data layer to the BizOps side. What were the biggest challenges your new team faced?

Our BizOps team has two primary goals: Looking at OKRs and KPIs to make sure they are being tracked correctly, and the second piece is working on the go-to-market strategy, pricing, and other higher-level work. The struggle operationally was that there were always so many moving parts, and so much of the work was manual. I knew we had to automate some of the processes.

How do you think about scaling an Operations team?

Operations teams often manage a lot of manual work that often goes untracked and is therefore difficult to quantify. For us, it was helpful to implement a tool like Avenue which lets us understand in more detail the activity of our operations team, how frequent certain problems occur, the SLAs of resolving them, and have a baseline to measure improvements over time.

I’m very excited for AngelList Stack. It has a lot of potential. When you think about AngelList Venture, there are heavy operations there. The Stack side is purely software, so it’s almost fully automated.

Documentation. When we started off, we had people who knew a lot of the process, but they didn’t have any of it documented. That person loses productivity because they are helping everyone, and it becomes a giant catch-up process.

I suggest stress-testing your documentation by giving a playbook to someone who doesn’t know the process and is seeing the guide for the first time. Then you’ll really know if you have documentation.

Finally, having some sort of cadence or check-in to make sure all the documentation is up-to-date and that the mechanics of the operation are still going smoothly is the final piece.

This post is part of Avenue's Fireside Chat Series.

Avenue sits down with AngelList Venture's BizOps Lead, Nigel Koh, to discuss all things data and ops. 

Fireside Chat with AngelList Venture’s BizOps Lead Nigel Koh

 is a social commerce food company founded by a team from Stanford. Jupiter helps food creators from TikTok & Instagram launch their own grocery powered recipe shops to sell directly to their followers nationally (similar to Shopify). Using Jupiter, you can meal-plan directly from your favorite food creator. Jupiter grew to $2.1M revenue in the Bay Area, before expanding to over 200M people in the U.S early this year.

The vision for the team at Jupiter is that in the not-so-distant future, every household's food will be managed by a system that functions like a personal shopper. A part of that experience is integrating food shopping with reading popular food blogs or finding recipes and having the order for all the necessary ingredients at your fingertips.

, Jupiter Co-Founder & CEO, also runs the Product Team. "We're very data driven. When we deploy anything, we have a lot of metrics set up," he shared. Munroe had set up reporting tools to analyze new features, but realized he needed a more robust, customized system that would bring the predictive elements of Jupiter into the team's own tools and keep pace with the company's growth.

 meant getting insights into every edge of the product to the right person at the right time.

Munroe says that now he and his team wake up every day to a wide array of meaningful metrics and monitors, from how many users ordered using a new feature to how many unsubscribed. Avenue's alerts and reports inform his team when a feature has been successful and when it needs some additional work. "Avenue also tells us when there are bugs, and we can see if people aren't interacting with a given feature," Munroe explained.

Jupiter's Product Team depends on these alerts to determine exactly which features move the needle and what direction to take the product.

As Jupiter continues to roll out, they are building new ways to become embedded within their user's experience, from the initial phase of finding a recipe to making a meal. Data-driven decision making is guiding that journey. In automating his team's processes, Munroe is maximizing product data and using it to sprint ahead:

Product, Growth, and Operations teams are often building at the edge of a business. They need to solve for everything that may slip through the cracks, and often do so with sticks and glue stringing together internal dashboards and reports alongside Playbooks in Google Docs or Notion. They are fire-fighters that, 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, accountability, analysis, and escalations you need to confidently scale your team's operations, all in one place.

How Jupiter Drives Growth & Product Intuition with Avenue. Operational Analytics, Operations, Data, Product, Intuition, Growth.

How Jupiter Drives Growth & Product Intuition with Avenue

Thrilled to be recognized as an innovator in Operational Analytics as part of the Modern Data Stack Rocketship Awards 🚀

Here because of our wonderful customer, their support, and all the creative applications and use of a tool like Avenue.

Excited to continue to build the most useful tool to make Ops teams more proactive, data-powered, and effective.

Avenue Operational Analytics Modern Data Stack Award

At Avenue we empower Operations teams to direct their attention most effectively. Today, we have launched support to set up muting rules so teams can mute unwanted alerts, reduce any noise and focus on the issues that really matter.

Automated alerts can sometimes be noisy. If there is a known issue that Operations teams do not need to be notified of, automatic alerts triggered as a result of this issue are added noise and can distract from more critical, unknown issues. Now, with muting rules, teams have the power to prevent unnecessary alerting and filter out expected events on Avenue.

Teams will be able to define muting rules on an alert’s signal details page. They must select the value from the alert’s available dataset they would like to mute. They can also specify until when they’d like to mute this alert. Operation teams can use any alert, including past alerts, to set up muting rules immediately!

Ability to set up rules that will mute specified alerts.

You can set up a muting rule for an alert on its signal details page. You will need to specify the time period this muting rule will be active and also which value from the alert’s data you would like to mute on.

How can I see all alerts that have been muted?

For a monitor, all muted alerts can be found on the monitor details page under Muted Signals. From this list, you can also click on any of the alerts to navigate to the details page and see why it was muted.

On the monitor details page, you can also see all muting rules currently defined for that monitor in the Mute List.

You can unmute an alert from the alert’s details page — this will remove the muting rule associated with that alert. As a result, you will also unmute all other alerts that have been previously muted by this rule.

Latest posts

How Avenue Makes Effective Logging Effortless with OpenTelemetry

Avenue Launches Monitor Tags

Avenue Launches Custom Actions in Zendesk

Independence for Operations Teams: A Customer Case Study with Mia Share

Fireside Chat with Mia Share's Director of Ops Molly Roberson

Avenue Launches Custom Actions in Customer.io

Fireside Chat with Jupiter's CEO & Co-Founder Chad Munroe

Avenue Adds Monitor Grouping and Bulk Actions

How Jupiter Drives Growth & Product Intuition with Avenue

Avenue Adds Role Based Access

Fireside Chat with Vori's Head of Ops Maika Koehl

How Vori Scaled Operations with Avenue

Fireside Chat with AngelList Venture’s BizOps Lead Nigel Koh

Avenue Operational Analytics Modern Data Stack Award

Avenue Launches Mute Lists

Avenue Launches Escalations with Message

Launch Week 2022 | Alerting Conditions + Filters

Launch Week 2022 | Data Views

Launch Week 2022 | Reporting, Grouping, + Signals

Fireside Chat with Blank Street's Head of Data, Ali Salihoglu

Avenue Launches Control Center for Ops

Avenue Launches In-App SQL Editor

Avenue Launches Advanced Issue Details: All Context and Action in One Place

Introducing Avenue: The First Operations Observability Platform