Trunk extends its developer toolkit with CI analytics

Trunk, a startup that aims to build a toolkit that helps developers build and ship their code faster, today announced the launch of its latest product: CI Analytics. The new service helps developers understand how their CI Workflows (currently with a focus on GitHub Actions) perform in the real world — and if there are any trends they should be aware of.

Founded in 2021 by a group of former Uber engineers, Trunk already offers Trunk Check, a tool for checking code quality, and Trunk Merge, a service that orchestrates merging pull requests. With CI Analytics, it’s now expanding this feature set with another tool that tries to help developers work more efficiently.

Image Credits: Trunk

“I’d run these surveys and the number one issue coming back from folks is ‘the hardest part of my job is landing code and merging code into main.’ That’s insane. We’re trying to build future-forward tech to make cars drive themselves and the hardest thing for the engineers is to put their code into the codebase,” Trunk co-founder and co-CEO Eli Schleifer told me of his time at Uber. “Every company has to invest a tremendous amount of money into this stuff and you really don’t want to hire 30 engineers — that’s how many were at Uber — to build this solution, because it’s not germane to your problem space. It’s just the core tax you pay.”

Schleifer described the new analytics service as an “engineering intelligence solution” that helps developers fix broken engineering workflows. He noted that while GitHub Actions has become very popular in a short amount of time, it’s also a bit of a black box. “There are a lot of engineering intelligence tools out there that will tell you that this engineer wrote this many lines of code or this many commits. We see engineering intelligence more as a tool to help the productivity of all the engineers,” Schleifer said. If Trunk can help these engineers find inefficiencies in their CI processes, then, he argues, it will make everybody in the engineering organization more efficient.

“Without a proper engineering intelligence solution, DevOps and engineering teams are left operating in the dark and engineers are left to guess at what parts of their build and test workflows are slowing down engineering,” said Schleifer. “Trunk CI Analytics eliminates the guesswork with beautiful trend lines, anomaly alerting, and the ability to perform deep historical analysis within a few clicks. Operating without this level of engineering intelligence can be the difference between shipping code on time and slowly grinding to a halt.”

The new service is now available to all Trunk users, with pricing starting at $7 per month and user (after a free two-week trial).

Trunk extends its developer toolkit with CI analytics by Frederic Lardinois originally published on TechCrunch

CircleCI partners with GitLab

CircleCI, the popular continuous integration and delivery (CI/CD) platform, today announced a partnership with GitLab, the popular DevOps platform that also offers an integrated CI/CD service. That may seem like a bit of an odd match-up at first, but CircleCI argues that by providing support for GitLab SaaS customers, it can help those users manage software changes anywhere in the software ecosystem.

Specifically, this means developers can now trigger pipelines in CircleCI from GitLab forked merge requests, for example, and CircleCI will be able to provide performance and health metrics to its customers on GitLab SaaS, all while also benefitting from CircleCI’s permission models and existing toolchains.

“When we launched 11 years ago, the only flow that people really cared about was: a developer makes a change in code, they commit, they push. That was all that existed,” CircleCI CEO Jim Rose said. “And over the last 11 years, it’s just metastasized into all these other things.”

circleCI and GitLab partnership

Image Credits: CircleCI

He acknowledged that for some users, simply using something like GitLab’s own end-to-end solution will work just fine for them. “That’s great for them,” he said. “I think in most situations, what we have found is that due to the complexity of the number of [version control systems] and due to the complexity of the number of deploy targets — most people we talk to are multi-cloud — you end up with a bit of a mishmash in the middle. And so we find our sweet spot is if you’re running multiple [version control systems] and multiple sources of change and you’re running multi-cloud, you may not benefit from attaching your build system to only one of those endpoints and then trying to stretch it out. Instead, what we try to do is go in as an agnostic broker and that allows you to make some of these different decisions individually.”

For CircleCI, it’s this focus on being the one service to manage code changes, something that becomes even more important as we’re moving to a world where an increasingly large part of a company’s infrastructure is managed by ‘[x] as code’ services, that allows it to occupy this neutral position where it can both partner and compete with the likes of GitLab, GitHub and Atlassian.

GitLab expands its DevOps platform with new observability and security tools

GitLab, the popular open source DevOps service, today announced version 15 of its platform. As usual, GitLab is basically announcing its roadmap for the entirety of the GitLab 15 release cycle here, with all of the new features rolling out over the course of the next year or so. The company says it will focus on observability, security, enterprise planning and workflow automation.

One of the first new features GitLab plans to launch is based on its acquisition of Opstrace last December.

“We’re going to continue to integrate not just an observability platform — available by default for developers to store their observability data and easily instrument their apps — but bring really great experiences that no other DevOps platform can create by showcasing that information in the right context for developers,” Kenny Johnston, product leader at GitLab, told me during an interview at last week’s KubeCon Europe event. “Think about things like developers writing code and being aware of recent incidents or errors related to that specific part of the code.”

Image Credits: GitLab

But in addition to giving developers more context around how their code performs (or fails), the new release will also give teams more insights into their overall DevOps process. Since teams are doing all of their DevOps work on GitLab, the service can provide metrics like time to merge, change failure rates or deployment frequency. GitLab is now expanding its support for all DevOps Research and Assessment (DORA4) metrics. “What we’re realizing is that it’s not just one number,” Johnston said. “It’s also important how you slice that number, so you need that number by an individual team, for example. [ … ] You can get a kind of team-based score for each one of those [teams] and then organizations can say: What’s that team doing that these other teams should be doing?”

On the security side, GitLab is adding new software supply chain security features in this cycle (no surprise there, given that it’s maybe the hottest topic in security right now). Teams will soon be able to automatically generate a Software Bill of Materials with a signed attestation for build artifacts, for example. Security teams will also be able to set group-level security policies soon, and the company is adding next-gen security testing tools to its scanners.

Also new in this release are additional features in GitLab’s planning tools to provide teams with more flexibility as to which frameworks they want to use, as well as a few quality-of-life changes like the addition of saves views and queries. GitLab 15 will also get the ability to automatically select code reviewers and the next workflow steps.

More importantly, though, GitLab is also launching new tools to make it easier for developers and the data science teams that are building ML models to work together. “We’re seeing this very common pattern of a mix between application, software data and models — and the DevOps process for that is that we write new code that’s going to take advantage of this new model. But we have a really hard time ensuring that the changes we’re making in the model and the code are combined together because the model is in a different application than where the code is.”

Remote raises $300M more, now at a $3B+ valuation, to manage payments and more for globally distributed workforces

Remote workforces have come into their own in the last several years, with companies ever more willing to tap into talent wherever it happens to be, and a vast array of low-friction tools being built to make those distributed teams work just as effectively as if they were all in the same physical space. Today, Remote, which has built a platform to hire distributed employees, and then make sure they are remunerated easily and legally — in other words, tech that helps companies with some of the trickiest aspects of managing a remote workforce — is announcing a big round of funding as it emerges as one of the bigger players to watch in the world of HR addressing global and distributed workforces.

The startup has raised $300 million, funding that it will be using to continue building out the tools that it provides to its customers and to expand its technology and services to more geographies. SoftBank Vision Fund 2 is leading the round, with previous investors Accel, Sequoia, Index Ventures, Two Sigma Ventures, General Catalyst, 9Yards, Adams Street and Base Growth also participating. This Series C values Remote at over $3 billion.

The size of the funding round and its timing — it’s been less than a year since Remote’s Series B, a $150 million round at a $1 billion+ valuation in July 2021 underscores a couple of things. One is the focus distributed work has had particularly in the last couple years — a trend that was already in pace before Covid-19 but definitely accelerated as a result of it; two is how Remote itself has expanded in that time.

The company — based out of San Francisco but with a totally remote workforce itself, with its two co-founders based in Europe — says that the number of employees processed through the Remote platform grew by 900% in the last year, with revenues up 13x in the same period (we have asked and the company does not disclose actual revenues or other specific numbers). That pace does not appear to be slowing down, even as offices gradually reopen and many parts of the world look to return to pre-Covid routines.

At the other end of the tech world spectrum, there’s been evidence that some of the funding exuberance of the last couple of years around pandemic-spurred theses (like rising demand in categories like remote work and delivery) is getting more bearish. But that trend too appears to have passed over Remote, which raised this round in the last quarter.

“The power dynamics have completely changed between employers and employees,” Remote CEO Job van der Voort said in an interview, with people more empowered he noted to work from wherever they want, and companies needing to provide remote working facilities to secure the talent at the price they want. “We only see this accelerating. If there were a slow down in that trend, maybe we couldn’t have raised this much.”

Remote’s customers now range from small startups to large enterprises and includes GitLab, DoorDash, Hello Fresh, Loom and Paystack, with companies sometimes processing payments and more for as few as four employees through Remote, while others are processing for thousands. Services it offers today include payroll, benefits, taxes and local compliance (including Employer of Record services) for contractors and full-time employees.

As for its footprint, currently, Remote says it provides services to “over 60” countries, but Job van der Voort, the CEO who co-founded Remote with Marcelo Lebre (COO and CTO), said that the aim to expand that to 100 this year, ultimately serving 140 countries.

The challenge that Remote is addressing is longstanding in the world of work, one that has been exacerbated with globalization. Hiring and then managing the administration of contractor or employed hires — when they are not based out of a company’s main office and country, and potentially not in any office at all but at home — can be a thorny business, crossing a number of different challenges in areas like international banking, local labor regulations and human resources management. Typically, companies have addressed this by working with local employment companies who have handled various processes manually for them, which led to an expensive and fragmented approach that ultimately held companies back from wanting to embark on the process at all; or not following policies that would be more beneficial for the company and its workers in the long run.

Van der Voort, who had previously been VP of product at GitLab, where he was a supporter of remote work but also someone who understood those challenges first-hand: he helped to build that organization’s remote team to 450 employees from just five. Lebre, meanwhile, had been the VP of engineering for Unbabel, which builds tools for companies to communicate with a global customer base, where he too worked with a distributed team and also saw the opportunity of addressing this area in a better way.

There are a number of tech startups in the market today that are tackling different aspects of remote employment, including the likes of Papaya Global, Oyster, Deel, HackerRank, and Turing. Remote’s unique selling point has been to build its stack from the ground up, building and providing Employer of Record services, fully operational legal entities, payroll and benefits, visa and immigration support and employee relocation, all provided in the cloud so that an employer can manage teams in different places from a single dashboard.

The company’s pace of growth in terms of its footprint speaks not just to the complexities and challenges of building out services like these, but also that integrated approach that Remote has taken in doing so.

“The reality is that it’s very difficult to open a new country and sometimes the reasons for a delay are out of our control,” Van der Voort said.

The integrated approach speaks to the tech chops of the company and how it will scale. Notably, Papaya Global made an acquisition of Azimo the other week specifically to bring money transfer services into its own fold — a feature that Van der Voort noted Remote already had in its stack.

“The way people work has permanently changed and the shift to remote and hybrid work has enabled companies to hire from anywhere in the world, but this can be an intensive, costly and risky process”, said Brett Rochkind, managing partner at SoftBank Investment Advisers, in a statement. “Remote has built a full-stack, global platform that creates a fast, seamless experience to hire and onboard new employees regardless of where they are. We are excited to partner with Job, Marcelo and the team to support their mission to open up the vast potential of the world for every person, business and country.”

 

Docker founder launches Dagger, a new DevOps platform

It’s been almost exactly four years since Docker founder Solomon Hykes left the company that kickstarted the container revolution. Docker has gone through its share of ups and downs since then, including selling its enterprise business to Mirantis in 2019, but Hykes, who was long the public face of Docker, mostly stayed on the periphery, with the exception of his participation in a few funding rounds. For a while now, though, he’s been quietly working on his next startup, Dagger, which is launching into public beta today and announcing a $20 million Series A funding round.

The round was led by Redpoint Ventures, with participation from Y Combinator, Nat Friedman (former CEO, GitHub), Brian Stevens (former CTO, Google Cloud and former CTO, Red Hat), Idit Levine (founder and CEO, solo.io), Julius Volz (creator of Prometheus), Ellen Pao (former CEO, Reddit) and Daniel Lopez (co-founder, Bitnami). Previously, Dagger raised a $3 million pre-seed and $7 million seed round led by New Wave.

Dagger, which was co-founded by Hykes and his fellow Docker alums Sam Alba and Andrea Luzzardi, aims to build what the team calls a “devops operating system.” Hykes noted how starting this new venture began with the team and not necessarily a product idea. The co-founders went looking for problems they could solve for the developer community and it quickly became clear to them that the DevOps process remains a bottleneck.

“We decided to start from zero and not assume we know anything,” Hykes told me about the process the team used to develop its ideas. “We began this long discovery process to just be a blank slate and listen to people’s problems. And they pulled us very quickly in the direction of CI/CD and automation pipelines. You know, you’ve got dev — and developers are happy and productive. You’ve got ops — things scale, there’s all this cool cloud stuff — and the glue in the middle, the DevOps part, that’s just really complicated. People find a way, but they just don’t like the experience and they all waste time and resources doing it. So we focused on that.”

The team argues that there are a lot of very powerful DevOps tools, but they tend to be very specialized — and as these applications expand in scope, so does the DevOps stack. “There’s no shortage of specialized tools, but then [developers] have to glue them all together — and the glue is the bottleneck. So we’re focusing on replacing the glue with something better,” said Hykes.

Specifically, that means Dagger lets DevOps engineers write their pipelines as declarative models in CUE (which stands for “configure, unify, execute”). With this, engineers can describe their pipelines and connect the different pieces to each other, all in code. Dagger calls these individual pieces “actions” and they, too, are described declaratively.

“The main difference is that it’s basically a real software development experience,” Hykes explained. “So if there’s an action that you like that someone else wrote, you can import it. If you want to look at the source code of that action, you can look at it, it’s in the same language you already know. And that action probably is built by combining smaller and more specialized actions all the way down. So it’s more like regular software.”

To complement the developer experience, the Dagger team is also building what it calls “Dagger Universe,” a curated library of packages that developers can import into their Dagger configurations.

This overall approach also means that potential users can keep their existing CI infrastructure in place. Dagger isn’t meant to be a replacement for the likes of Circle CI or GitLab — it’s basically a layer on top of that.

“It is far too complicated for DevOps teams to manage their infrastructure and deploy software to different clouds, but Dagger has elegantly cracked the code to streamline software supply chain management,” said Erica Brescia of Redpoint Ventures. “By making custom app delivery pipelines portable, the Dagger team has changed the game for building and deploying software.”

Hykes noted that he is taking quite a few learnings from his experience at Docker into building Dagger. Like Docker, Dagger will have an open-source piece to it and while the team is still figuring out the details, it’s going to be a key part of the Dagger ecosystem.

“Dagger is going to be a hybrid platform,” Hykes explained. “So there’s an open source engine, which we’re launching [today], and there will be an optional cloud service that will be very tightly integrated, but still optional. […] Our conclusion from Docker is, if you want a large and thriving developer community, you need a real open source project. It can’t be fake open source. But if you want that community to continue thriving — and if user experience specifically is important — then you need to connect that community to one product vision and not 10,000 different product visions.”

For now, the team is going to focus on this open source engine to see what the community needs and where the pain points are. The managed service will come later. Hykes noted that at Docker, everything happend so quickly and the service became such a foundational technology almost overnight, the company was pulled into too many different directions. With Dagger, he plans to take things slower — and because Dagger doesn’t run the applications itself, he believes that the team will be able to maintain this focus.

“We’re going to do the same thing with commercialization. I think with commercialization, at Docker, we felt like there was a playbook that we were obligated to follow and we didn’t really listen to our community enough,” Hykes said.

Dagger is going to use the new funding to expand its engineering team to build out its product, but the company is also hiring to build out a marketing and developer relations team.

CircleCI acquires Ponicode

Continuous integration and delivery platform CircleCI today announced that it has acquired Ponicode, a Paris, France-based startup that helps developers automate their unit tests and track their overall test coverage. Ponicode, which was founded in 2019 and had previous raised $4.6 million in funding, will continue to operate as usual for the time being, but

With this move, CircleCI aims to get closer to the developer workflow. While the company already offers some testing capabilities, its typical user base is operations teams that set up and manage their CI/CD pipelines through its tools. The company wants to get closer to the developers, though, and with Jean-Philippe Leblanc, it recently hired a new senior VP of engineering to help it do so.

Image Credits: CircleCI

“Shifting left in front of developers, that’s something we always want to do,” he told me. “Even though we are a developer tools [company], per se, it’s more at the organizational level. For us to get ourselves more in front of developers is actually key.”

This also means that while CircleCI plans to bring some Ponicode capabilities into its own testing platform and hopes that its 16-person team can bring its machine learning and testing expertise to the company as a whole, Ponicode will remain a tool that will continue to work with a large variety of CI/CD platforms that compete with CircleCI. Leblanc noted that the focus here is on offering value at the developer level, no matter whether that developer also uses other parts of its platform. “Our biggest wish is actually that if you’re a developer and you have the CircleCI/Ponicode plugin installed and you move to Company B — even though they’re using Travis or Jenkins or whatever, you’re still installing the CircleCI/Ponicode plugin. That would be ‘mission accomplished’ from our perspective.”

He also noted that it will take a while to figure out how to best integrate Ponicode into CircleCI’s overall testing product. “For us, it’s going to be an integration that goes both ways. How do we inject ourselves — CircleCI into the IDE, terms of our CLI or APIs,” he  said. “And there is build status. Can we eventually start doing those continuous micro-validations for the developer before it actually reaches somewhere else? Making sure not just the code but all the dependencies check out.”

Ponicode’s AI engine allows developers to automatically ensure test coverage for their code — and ideally find issues before that code even hits the continuous integration platform and its testing services. In addition, Ponicode also offers developers a dashboard for keeping track of their test coverage and other metrics. That’s something the CircleCI team is thinking to integrate into its CI/CD pipeline, making it easier to scan an entire repository and providing organizations with better data about how their codebases are covered.

“The daily work for a developer is filled with obstacles that keep them from writing code, such as creating tests, debugging issues, responding to production issues, and so on,” said Ponicode CEO and co-founder Patrick Joubert. “These problems continue to increase for developers as more and more organizations look to ‘shift left.’ We’re confident that the power of Ponicode’s proprietary AI engine combined with CircleCI’s CI/CD platform will further aid developers to write better code.”

CircleCI acquires Ponicode

Continuous integration and delivery platform CircleCI today announced that it has acquired Ponicode, a Paris, France-based startup that helps developers automate their unit tests and track their overall test coverage. Ponicode, which was founded in 2019 and had previous raised $4.6 million in funding, will continue to operate as usual for the time being, but

With this move, CircleCI aims to get closer to the developer workflow. While the company already offers some testing capabilities, its typical user base is operations teams that set up and manage their CI/CD pipelines through its tools. The company wants to get closer to the developers, though, and with Jean-Philippe Leblanc, it recently hired a new senior VP of engineering to help it do so.

Image Credits: CircleCI

“Shifting left in front of developers, that’s something we always want to do,” he told me. “Even though we are a developer tools [company], per se, it’s more at the organizational level. For us to get ourselves more in front of developers is actually key.”

This also means that while CircleCI plans to bring some Ponicode capabilities into its own testing platform and hopes that its 16-person team can bring its machine learning and testing expertise to the company as a whole, Ponicode will remain a tool that will continue to work with a large variety of CI/CD platforms that compete with CircleCI. Leblanc noted that the focus here is on offering value at the developer level, no matter whether that developer also uses other parts of its platform. “Our biggest wish is actually that if you’re a developer and you have the CircleCI/Ponicode plugin installed and you move to Company B — even though they’re using Travis or Jenkins or whatever, you’re still installing the CircleCI/Ponicode plugin. That would be ‘mission accomplished’ from our perspective.”

He also noted that it will take a while to figure out how to best integrate Ponicode into CircleCI’s overall testing product. “For us, it’s going to be an integration that goes both ways. How do we inject ourselves — CircleCI into the IDE, terms of our CLI or APIs,” he  said. “And there is build status. Can we eventually start doing those continuous micro-validations for the developer before it actually reaches somewhere else? Making sure not just the code but all the dependencies check out.”

Ponicode’s AI engine allows developers to automatically ensure test coverage for their code — and ideally find issues before that code even hits the continuous integration platform and its testing services. In addition, Ponicode also offers developers a dashboard for keeping track of their test coverage and other metrics. That’s something the CircleCI team is thinking to integrate into its CI/CD pipeline, making it easier to scan an entire repository and providing organizations with better data about how their codebases are covered.

“The daily work for a developer is filled with obstacles that keep them from writing code, such as creating tests, debugging issues, responding to production issues, and so on,” said Ponicode CEO and co-founder Patrick Joubert. “These problems continue to increase for developers as more and more organizations look to ‘shift left.’ We’re confident that the power of Ponicode’s proprietary AI engine combined with CircleCI’s CI/CD platform will further aid developers to write better code.”

Propelo raises $12M for its AI-driven engineering excellence platform

The number of DevOps tools has increased exponentially over the last few years and, with that, the amount of data these tools can produce to help businesses improve their software development processes. But most of the time, this data simply ends up in a dashboard for some scattershot analysis. Propelo (previously known as LevelOps) wants to bring order to this chaos and aims to build an “AI-driven engineering excellence platform” that brings together a set of machine learning-powered analytics services and no-code robotic process automation (RPA) tools to help users turn these data points into something actionable.

The company today announced that it has raised a $12 million Series A funding round led by Decibel Partners. Fike Ventures, Eniac Ventures and Fathom Capital also participated in this round.

Propelo founder and CEO Nishant Doshi previously co-founded SaaS security service CirroSecure, which Palo Alto Networks acquired in 2015. After that, he spent a few years at Palo Alto, where he experienced the explosion of DevOps tools firsthand as a senior director and VP of engineering. To get a better view of the development process, the team had to stitch together data from sources like Jira, GitHub, Salesforce and elsewhere.

Image Credits: Propelo

“That’s very manual, very resource-intensive,” he said. “You’re not focusing on the core of your businesses but you start hacking together a solution — and there’s always one more tool. And even if you get these tools, you don’t know what to measure. You won’t have access to the advances that a purpose-built solution like ours brings to the table, and more importantly, you don’t have actionability.”

Image Credits: Propelo

It’s this last part, Doshi stressed, that closes the circle. It’s one thing to have good data and analytical insights, after all, but that’s not going to help your development processes if there is a disconnect between this information and actually taking actions based on it. Using Propelo’s RPA tools, users — and the company notes that it mostly caters to users in the engineering leadership stack — can easily automate a lot of tasks and workflows to improve DevOps processes within a business.

The service currently integrates with about 40 DevOps tools like Jira, GitHub, GitLab, Jenkins, Gerrit and TestRails. Using its AI smarts, Propelo can then help its users discover hidden bottlenecks or predict when a sprint will likely slip — which is difficult, because developers have a tendency to never move a task from “in progress” to “done.” Indeed, data hygiene — and updating Jira tickets — isn’t something most developers really want to think about, so Propelo can regularly nudge them to do so.

Current Propelo users include the likes of Broadcom and CDK Global. “Propelo provides us the data-driven insights on how to reduce DevOps friction and reduce wasted motions at a very granular per scrum team level. This helps us with maximizing the efficiency of additional engineering investments and removing pain points for engineers,” said Joe Chen, a VP of Engineering at Broadcom in charge of security technologies and endpoint solutions.

A startup’s guide to software delivery

One of the biggest factors in the success of a startup is its ability to quickly and confidently deliver software. As more consumers interact with businesses through a digital interface and more products embrace those interfaces as the opportunity to differentiate, speed and agility are paramount. It’s what makes or breaks a company.

As your startup grows, it’s important that your software delivery strategy evolves with you. Your software processes and tool choices will naturally change as you scale, but optimizing too early or letting them grow without a clear vision of where you’re going can cost you precious time and agility. I’ve seen how the right choices can pay huge dividends — and how the wrong choices can lead to time-consuming problems that could have been avoided.

The key to success is consistency. Create a standard, then apply it to all delivery pipelines.

As we know from Conway’s law, your software architecture and your organizational structure are deeply linked. It turns out that how you deliver is greatly impacted by both organizational structure and architecture. This is true at every stage of a startup but even more important in relation to how startups go through rapid growth. Software delivery on a team of two people is vastly different from software delivery on a team of 200.

Decisions you make at key growth inflection points can set you up for either turbocharged growth or mounting roadblocks.

Founding stage: Keep it simple

The founding phase is the exciting exploratory phase. You have an idea and a few engineers.

The key during this phase is to keep the architecture and tooling as simple and flexible as possible. Building a company is all about execution, so get the tools you need to execute consistently and put the rest on hold.

One place you can invest without overdoing it is in continuous integration and continuous deployment (CI/CD). CI/CD enables developer teams to get feedback fast, learn from it, and deliver code changes quickly and reliably. While you’re trying to find product-market fit, learning fast is the name of the game. When systems start to become more complex, you’ll have the practices and tooling in place to handle them easily. By not having the ability to learn and adapt quickly, you give your competitors a massive edge.

One other place where early, simple investments really pay off is in operability. You want the simplest possible codebase: probably a monolith and a basic deploy. But if you don’t have some basic tools for observability, each user issue is going to take orders of magnitude longer than necessary to track down. That’s time you could be using to advance your feature set.

Your implementation here may be some placeholders with simple approaches. But those placeholders will force you to design effectively so that you can enhance later without massive rewrites.

Very early stage: Maintain efficiency and productivity

At 10 to 20 engineers, you likely don’t have a person dedicated to developer efficiency or tooling. Company priorities are still shifting, and although it may feel cumbersome for your team to be working as a single team, keep at it. Look for more fluid ways of creating independent workstreams without concrete team definitions or deep specialization. Your team will benefit from having everyone responsible for creating tools, processes and code rather than relying on a single person. In the long run, it will help foster efficiency and productivity.

A startup’s guide to software delivery

One of the biggest factors in the success of a startup is its ability to quickly and confidently deliver software. As more consumers interact with businesses through a digital interface and more products embrace those interfaces as the opportunity to differentiate, speed and agility are paramount. It’s what makes or breaks a company.

As your startup grows, it’s important that your software delivery strategy evolves with you. Your software processes and tool choices will naturally change as you scale, but optimizing too early or letting them grow without a clear vision of where you’re going can cost you precious time and agility. I’ve seen how the right choices can pay huge dividends — and how the wrong choices can lead to time-consuming problems that could have been avoided.

The key to success is consistency. Create a standard, then apply it to all delivery pipelines.

As we know from Conway’s law, your software architecture and your organizational structure are deeply linked. It turns out that how you deliver is greatly impacted by both organizational structure and architecture. This is true at every stage of a startup but even more important in relation to how startups go through rapid growth. Software delivery on a team of two people is vastly different from software delivery on a team of 200.

Decisions you make at key growth inflection points can set you up for either turbocharged growth or mounting roadblocks.

Founding stage: Keep it simple

The founding phase is the exciting exploratory phase. You have an idea and a few engineers.

The key during this phase is to keep the architecture and tooling as simple and flexible as possible. Building a company is all about execution, so get the tools you need to execute consistently and put the rest on hold.

One place you can invest without overdoing it is in continuous integration and continuous deployment (CI/CD). CI/CD enables developer teams to get feedback fast, learn from it, and deliver code changes quickly and reliably. While you’re trying to find product-market fit, learning fast is the name of the game. When systems start to become more complex, you’ll have the practices and tooling in place to handle them easily. By not having the ability to learn and adapt quickly, you give your competitors a massive edge.

One other place where early, simple investments really pay off is in operability. You want the simplest possible codebase: probably a monolith and a basic deploy. But if you don’t have some basic tools for observability, each user issue is going to take orders of magnitude longer than necessary to track down. That’s time you could be using to advance your feature set.

Your implementation here may be some placeholders with simple approaches. But those placeholders will force you to design effectively so that you can enhance later without massive rewrites.

Very early stage: Maintain efficiency and productivity

At 10 to 20 engineers, you likely don’t have a person dedicated to developer efficiency or tooling. Company priorities are still shifting, and although it may feel cumbersome for your team to be working as a single team, keep at it. Look for more fluid ways of creating independent workstreams without concrete team definitions or deep specialization. Your team will benefit from having everyone responsible for creating tools, processes and code rather than relying on a single person. In the long run, it will help foster efficiency and productivity.