For Tina Huang, the idea of automation without human intervention is wrong. That comes from his studies at the University of Chicago in cultural anthropology and blogging, as well as stints on Google and Twitter.
“I worked at Google News, and then, later, switched to Twitter. And I had this tendency around people in the loop, right? – like how do you add artificial intelligence to human editorial?” said Huang, found and Chief Technology Officer at Transposit.
The San Francisco -based startup is focused on integrating the tools used in DevOps and site reliability engineering (SRE). It aims to provide a source of truth to calm chaos through automation and human action.
He argues that most workflow tools, such as Zapier and others, seek full automation without human intervention.
“That’s really nice,” he said, “but it’s very limited in the kind of problems that can help.”
People often suggest that he will lose his job if people completely automate everything. His response:
“Well, that’s true if you have a system in a loop with no change on the inside and no change on the outside.”
An incident may occur with the change: I’m adding new features to my codebase. I’m adding new infrastructure to my stack. But changes can also happen on the outside.
“As with Twitter, a lot of incidents happen because you have a different traffic pattern because of the Super Bowl, or some event that you didn’t expect.
“I remember the Arab Spring bringing challenges to Twitter because that was like a spike in usage and a different style of usage than we expected,” he said.
Made in SQL
Huang and Adam Leventhal founded the company in 2016, having started Divanny Lamas from Splunk as CEO. Leventhal and Lamas returned to Sutter Hill Ventures, one of its funding partners. The company raised $ 50.4 million in total after a $ 35 million Series B last year led by Altimeter Capital Management.
Huang saw the world change for developers, instead of building something from scratch, it was becoming more common to integrate components from different data sources connected by APIs.
They spent three years creating a platform to combine all the workflow tools in one place. In their Series A pitches, they described it as a relational database of APIs.
“Imagine that you could just query your data with the ease of something like SQL, and not worry if that data comes from GitHub, that comes from Slack, that comes from any of the different these resources? ” said Huang.
Looking at technology, they realize that DevOps and IT operations are the most likely fit for the idea of human-in-loop response, even while directing machine learning to the tasks at hand.
They likened their technology to SQL, which gave developers a way to discuss what data they were trying to retrieve without thinking about how to actually access the data:
“At the very least, that’s the basis of Transposit integration technology. So you can write an SQL query, and we’ll deal with authentication, pagination, type of throttling of queries, etc.…
“One of those areas that sets us apart from almost any integration platform I’ve seen there is the depth to which we’ve taken that where it’s not just a thin wrapper over API calls. But it really allows the developer to not think about the underlying services. ”
Automated Documentation
They applied their ideas initially to runbooks for incident management, the binder some companies made for 2 am hair’s-on-fire calls when something broke. In those instances, the on-call engineer may be sweating over details of code he didn’t write or operational procedures he wasn’t familiar with.
The Transposit platform allows organizations to build automated runbooks that lay out a set of DevOps processes to be executed across multiple tools.
Huang admits that runbooks are not a sexy topic to follow, but insists that that’s largely because people have become uncreative with them – they’re basically flat wiki pages that everyone knows, or at least assumed, are not up to date, so they do not use them.
Transposit’s answer here is to make them more interactive and useful by combining semi-structured and unstructured machine data with human data.
You can set a trigger, which will start an automatic workflow of how to proceed. It could be things like starting a Slack channel, filing a ticket for it, assigning a commander, notifying the customer success that this incident is going on, knowing which team is involved and -page to the person being called.
“It could be everything from flat documentation to adding buttons that will help you, so you might have a button that says,‘ restart service, ’‘ update status page, ’etc.” says by Huang.
It can break loops and merge workflows between dev and ops with hundreds of pre-built integrations or easily connect to any API. Then, you can also have a cleaning process.
At the same time, it documents every step of the way, providing the full audit trail and postmortem report. Runbooks are not static documents. The underlying machine learning algorithms can look at how engineers recover and provide data-based revisions to processes. That will also gain knowledge of the institution.
In a blog post explaining the need for codified processes, both machine and human, Huang wrote:
“The ideal world is about being able to combine these two seamlessly. When you allow people to combine pieces of automation, you really have more automation and better automation.
“That’s because you can let human intuition cross over where the logic to automate would be so complex that it would cause more problems than it solves, where the logic would be closely tied to the details of a fast-evolving product that it will. needs to be updated almost as often as the product itself. “
Tickets Plus Action
Late last year, Transponsit announced Activities, which uses ticketing systems such as JIRA and ServiceNow.
The company does not call them tickets because it generally requires action beyond the ticket itself. It usually comes with a document, maybe a Confluence page, that tells you how you should use that JIRA ticket – for example, categorize the incident according to these circumstances. At the same time, people are talking about the issue with Slack or Microsoft Teams. There are repetitive actions required for filing any ticket-you need to fill in these 10 fields, perhaps.
“One of the great things about DevOps functionality is that many of them can also be automated,” Huang said. “And so are some pieces of it that require human intervention like the human router for transmission – we hear that a lot of people still use some form of NOC (network operations center).… Honestly, heard we’re that people outsource them to third parties.Their job is literally to look at some graph and determine what group the actual page is.right? So there’s a slight human element here.
“But once someone determines which team to reach, the rest of it can be automated.”
No-code Builder
The company treats Transposit as a code -free platform.
In a BetaNews article on 2022 predictions, Ed Sawma, Transposit’s vice president of operations, put:
Low-code and no-code technologies open up new paths to innovation by empowering more users in an organization-with or without technical experience-to contribute and add value to project or workflow. In 2022, we will see DevOps practices gaining wider use in functions outside of traditional software development. In addition to the growing use of code-free tools, non-technical users will experience all the benefits of a seamless delivery approach and will want to use it for everything digital they create.
The company has a no code platform as well as a more advanced developer platform. Its no-code builder can have a git repository associated with them. It is written in Python and designed to be very developer.
“What we saw from our first customers was [that] they rejoice in the fact that they no longer have to wait for us to do something for them. You don’t have to wait for us to build a data connector, ”said Huang.
“We are built on open standards in OpenAPI. So you know, if you have an internal system, anything API, you can plug into our system. … You can build almost anything you want. You know, we see customers coming together … transpose it to trigger AWS Lambda to call the edge nodes in their compute clusters. And so, it’s very strong and very, very flexible. ”