Workato

An SRE team from scratch for Workato: 10 senior engineers hired into a little-known company

Together with Workato we defined a role that had no description and hired 10 senior SREs into a new team

In 2020 Workato, a US enterprise automation company, needed a new team that had neither a role description nor even a name. Together with the client, NEWHR defined the role as SRE in Google's sense, searched among open-source contributors and hired 10 senior engineers. Three years later most of them were still with the company.

Request and constraints

In summer 2020, at the peak of the candidate market, Workato was a startup of a few dozen people (under 100). Today it is one of the leaders in enterprise automation: the company integrates systems and automates processes with AI, serves 17,000+ enterprise customers worldwide and has 1,000+ employees.

The task was to build an SRE team from scratch, at the intersection of systems programming and infrastructure. The search ran from July 2020 to March 2021; later NEWHR joined for individual requests.

  • Senior developers with 8+ years of experience and an engineering background
  • Understanding of computer science
  • Golang or Ruby
  • Infrastructure experience, deep knowledge of Kubernetes, Terraform, cloud
  • Experience with high-load, high-performance systems and various databases
  • At least 3 years at each company: the client did not consider job hoppers
  • Conversational English

Why it was hard

A new team with no role description

The new hires were to solve problems that did not match any familiar role. There were no clear tasks or job description yet.

A little-known company hiring only top-tier engineers

The company was almost unknown on the Russian market, and the role had to be "sold" to developers who received several inbound offers a week. The words "startup" and "option plan" tended to put them off.

Six interviews with the client

Candidates were interviewed in turn by 6 hiring managers with different expertise: theory, algorithms and data structures, live coding. In 2020 far from every developer agreed to such a marathon.

English

In 2020 few Russian developers were looking at international companies, so English was not a priority for them.

How we searched

The search was led by Ekaterina Veselkina, Head of Tech Recruitment NEWHR.

We brought a technical expert to the first briefing

The client came with a request for a whole team of "Distributed Systems Engineers" but did not know exactly how many people it needed, whether such candidates existed on the market or what the role was called at other companies. To understand the requirements correctly, NEWHR invited a technical expert to the first briefing: the expert asked in-depth questions about the tasks, and the recruiter moderated the conversation. This made NEWHR co-authors of the role, and together with the client we shaped the selection criteria.

The role turned out to be closest to SRE (Site Reliability Engineer) in Google's sense.

"In Russia at that time, SRE mostly meant system administrators. At Google, an SRE is a developer who automates tasks that would otherwise have to be done manually.

By the company's rules, SREs spend no more than 50% of their working time on operations (tickets, on-call, manual tasks). The other half should be devoted to development."

Ekaterina Veselkina, Head of Tech Recruitment NEWHR

We searched among open-source contributors on GitHub, not only on LinkedIn

Besides LinkedIn, Ekaterina searched on GitHub, among active contributors to projects such as Kubernetes and databases. The main "donor" companies were those with a serious multi-stage hiring process that build internal developer tools or are known for strong platform teams.

"I had two options: run volume interviews with candidates from LinkedIn and reject 9 out of 10, or go to GitHub and search among active contributors to specific projects, such as Kubernetes. Yes, there are fewer of them, but the hit rate will be 1 in 1."

Ekaterina Veselkina, Head of Tech Recruitment NEWHR

We sold the role through real engineering problems

To get busy developers to respond and agree to 6 hard interviews, Ekaterina focused on non-standard problems. She got examples of specific challenges the new team would face from the client and opened her messages to candidates with them: work at the intersection of infrastructure and systems programming, not CI/CD but incident analysis, finding bottlenecks, optimisation and architecture. For example, the team digs deep into Kubernetes, writes its own faster and more secure version of Docker, patches off-the-shelf software (Kibana, Kubernetes) for its needs and deals with memory leaks in Ruby.

We screened strictly before the client saw anyone

Out of hundreds of candidates reviewed, 22 CVs reached the client.

What Ekaterina looked at

  • The ability to switch between programming languages and choose the language for the task.
  • Understanding of computer science: good education and/or the nature of past tasks were the clues.
  • Previous employers: NEWHR knows which companies are known for strong infrastructure teams.

Who did not pass

  • Backend developers who worked mostly on product tasks, with no systems or platform development experience.
  • SREs who grew out of administrators rather than developers: they would not have passed live coding and algorithm tasks.
  • Job hoppers (the client's requirement).

We discussed hard technical problems from the first screening

From the first screening, Ekaterina discussed hard technical problems with candidates. This matched the client's culture: the company has several founding engineers, and they were looking for engineers as strong as themselves.

"I asked candidates 'not to spare me' and to describe their tasks in as much detail as possible. That way I could tell whether we meant the same thing when we said 'hard'.

I also told them that 6 two-hour interviews with our client were waiting for them, without trying to smooth things over. On the contrary, I even scared them a little: I only needed those who were ready for really tough challenges."

Ekaterina Veselkina, Head of Tech Recruitment NEWHR

We also checked soft skills: the client is a fast-growing startup with a fully distributed team, so it needed mature, independent and proactive engineers ready to own their decisions. English was checked separately: all work communication at the company is in English.

We agreed with the client on regular feedback

From the start, NEWHR agreed with the client to ask for feedback regularly, especially in the first weeks of the search. The client understood that hiring success depended on its involvement and gave detailed feedback on CVs and after every interview stage. This helped Ekaterina understand what to focus on next.

Ekaterina also spoke openly about her doubts on candidates and pointed out what to pay attention to in interviews: for example, when she was not sure a candidate had solved hard enough problems, or had doubts about their motivation.

"We built the most transparent and open communication with the client, and that was the key to our success. We were both interested in the best candidate reaching the final."

Ekaterina Veselkina, Head of Tech Recruitment NEWHR

Funnel

StageCandidatesConversion from previous stageConversion from start
Initial search603—100%
Agency interviews6310.4%10.4%
CVs sent to the client2234.9%3.6%
Client interviews22100%3.6%
Offers1359.1%2.2%
Started work1076.9%1.7%

Result

All 22 candidates NEWHR presented to the client were invited to interview. 13 received offers, and 10 joined the new SRE team.

Three years after the search, most of them were still working at Workato and were working with NEWHR as hiring managers.

"We have worked with NEWHR since 2018. NEWHR was the first agency we turned to and the only one that stayed… Many of the people they found have grown into Head-level roles."

Alexey Pikin, VP of Engineering, Workato

What this means for your hiring

Have a role nobody can fill? Let's talk.

contact@new.hr