A reversed interview by Olivier Dennemont with Antoine Craske, founder of the QE Unit, the Quality Engineering community. This interview lets Antoine share his personal story, background and perspectives on the software industry, and especially on the theme of Quality Engineering,
Organizations face tremendous pressure to reinvent their value proposition with software that meets the requirements of Quality at Speed. In such a context, companies are required to transform their software paradigm to QE for survival, hence the mantra “Quality Engineering or Die Trying”.
That mantra also reflects the motivations Antoine is having in driving the Quality Engineering community and messages within the industry,
Follow the QE Unit for more Quality Engineering from the community.
What’s your story and background?
Starting with the personal part, I was born in France, in the city of Reims. I like to recall it as the city of the Champagne that I enjoy from time to time. We stayed there for about 2 years with my twin brother, who also followed an engineering path but in the automotive sector.
At first I wanted to be a musician and eventually an orchestra maestro. I studied a lot of music, playing everyday for 1 to 2 hours for about 20 years. I had started with drums, percussions, to then focus exclusively on saxophone. I went to the conservatory in France after an initial failure lacking the right foundations and methodologies. But I trained again for 2 years and entered directly to get out with the diplomas, gold medals in solo and chamber music. Since a few years I kept piano as a learning leasure.
In a way I was “trained” to work and be busy all day. I know it is a personal driver for me to achieve concrete results everyday. I was going to music all Wednesday and Saturday afternoon with school in between, where it was an opportunity to build up my focus and discipline on top of creativity. I strongly recommend learning music for anyone, as we will see in software, we can learn a lot for other practices.
I was with an average rating at school as not finding practicality in what I was learning. I did not want more theory after the main graduation, not being interested to do preparatory classes for 2 years, to then go to an engineering school that was still general. I therefore choose a more practical course of 2 years called a “DUT” in France, aiming at being a junior software engineer. When I finished I also knew that, especially in France, schools count for your first opportunities, salary and perception.
I decided to continue with a license and master degree part-time with work to combine the requirements of a diploma while being able to work 3 years directly, instead of remaining at school. I was busy with 3-4 days a week at work and 1-2 days at school depending on the weeks; I stopped the music training at that time to dedicate myself to study and work. It was a very interesting degree called “MIAGE”, covering IT with a broad perspective from architecture, security, quality, integration, but also concrete practices like TDD or CI/CD. The course was completed with Business Administration, Finance & Management, giving a big picture and very complementary vision of IT for business.
At work I started as a software engineer at La Redoute in the international division, with the culture of a mid-size company, startup mindset, but with a big boat behind us with the French division. I remember the first day where I had a call with guys in the United States to discuss integrating web navigation data into a data warehouse. It was an interesting experience as I worked in data pipelines that we call data engineering nowadays, software engineering for other needs, and was also practicing small project management.
Curious and with more constraints of a small division and more international competition, we quickly adopted practices from the emerging world of “Continuous Integration” with jenkins pipeline, central logging, standard components per patterns, monitoring, etc. It was also in 2011 that I started to code an internal test automation framework to help accelerate software delivery and release. We used other open-source components like Selenium for testing the e-commerce website and CRM. In 2013, Cerberus Testing was born and open-sourced for the community. I am still involved now and the paradigm has evolved from a test framework to a test automation platform.
In that context I had the desire to evolve as project manager and got the opportunity to handle IT projects in a variety of themes like business projects, application modernization, infrastructures, integration, data centers, etc. I enjoyed getting that transversal perspective of IT and improving the efficiency of my project management. Driving projects was also a great way to understand inefficiencies in the end-to-end software lifecycle. I then got the opportunity to manage the international IT perimeter between 2013 and 2015, managing development, operations and data teams of about 50 persons on the entire scope from business roadmap to infrastructure management. I wanted to live abroad and took the step in 2015.
I moved to Portugal as I wanted to go to Portugal or Italy, always having a natural attraction in these two countries. I had a preference for Portugal with regular trips I had done and the opportunity was there. It was the transformation, or more precisely, drastic reinvention period for La Redoute in 2014-2019. The company was with 600m€ sales but with -50m€ EBITDA, without backup plans, and had 4 years to transform or close. I had a role of Transformation Program Director working on different aspects like migrating large application scope from a Mainframe to ERPs to internally developed applications, setting up a Datacenter in Russia, building up our Operations Center in Portugal, consolidating and rationalizing from multiple external partners, and keeping DevOps and CI/CD themes.
In 2017, I participated in a true Phoenix project, like the book. It was a very complex business transformation program for building one of the most automated warehouses in Europe to enable 2 hours order-to-ship. The program involved many streams like deploying and migrating to new management systems for stock, warehouse, order, transport for all countries, plus all integration work. I was appointed to lead the engineering teams in a very complex period where I worked 8 months without weekends and long hours, with the true help of a colleague of mine from France, Ali Hakim Benyahia, who also remained on-site, making the bridge with the board and product areas.
After that period, I evolved as an engineering manager with margin and autonomy to improve the system. The main problem was the fragmentation of the teams across multiple providers, a project organization instead of a product one, disconnected from the business flow of priorities and value delivery. I had a major stream of incorporating in Portugal all the engineering forces, removing step by step all partners and other resources in other structures. I worked on the organizational part, setting up cross-functional teams, structuring layers of managers, and also acting outside of my scope of responsibility to develop a product, devops culture. Reporting to the CIO, I got a broader and transversal picture on where to contribute and what to improve on the end-to-end lifecycle, thinking across silos and formal perimeters.
That experience led me to less than 2 years after to evolve as Senior Engineering Director to drive structuring changes transversally. I accelerated a lot on the DevOps paradigm using the framework Accelerate acting on the 4 metrics. One example was to iterate focusing on the software pipeline addressing different topics like methodologies of releasing, deployment, adapting the organization with responsibilities of software engineers up to production, technological ones like self-service templates, standard observability metrics or setting up the foundations of a modern cloud-native application development platform. I acted out of silos to bring value on the entire scope, making front, back, data and infrastructure engineering perimeters collaborate. I had the opportunity to share a Technology Strategy with the board on topics such as event-driven and streaming to deliver user experience and business operations improvements.
During that period where I got managerial responsibilities I participated a lot in meetups in communities in France and Portugal. I always had at least a meetup per week, and it was common to have multiple per week, mainly on-site at that time. I keep my vision on the software industry, participating in communities of software, devops, data science, data engineering, or even HR in technology. That’s where I got and I am still getting the good practices and the pains of the different verticals in software. I also started to share in conferences on topics we were addressing in the domains of DevOps, engineering, Quality Assurance and to animate community. I grew the TICE.Leiria meetup group to more than 800 members based in central Portugal, created a dedicated one on Kafka, another one for Cerberus Testing. More recently I also got involved with the Ministry of Testing animating the one of Leiria.
In July 2020, Jean-Cédric Costa arrived as the new CIO of La Redoute to support the growth period the company is now in. I evolved as Director of Architecture & Technology to drive the alignment from business objectives to the end-to-end IT alignment, and had to start from scratch the entire practice as that team was discontinued in the last years. We co-created with the business and are now driving a transformation plan with a 2025 perspective. Our target revolves around a business platform architecture to support the growth and scalability of the business in being the preferred lifestyle and family platform. This position is really interesting due to its end-to-end picture and transversality with Business, IT & Data. I have been involved in the business and leadership team since 2017, and that position had a very strong focus on that part and helped me to grow. Early this year in 2022 I took over the Cloud Center of Excellence perimeter to drive more directly the transformation and evolve with the responsibility of the “Technology Transformation”.
Today, I am still in Portugal. It’s the country where I have the feeling of being back at home when I travel.
What led you to work and define Quality Engineering?
I always enjoyed learning and understanding systems with the big picture, consolidating patterns and practices with the end-to-end perspective. My goal is to deliver results with efficiency, and I believe we need to understand a system before trying to improve it, whatever it is.
I improved my practices of software engineering, project management, management or even leadership over-time. After working on these silos, I was not satisfied with the results of my frameworks, with the feeling of missing something, and not having sufficient impact in the ecosystem. My experience was to align the entire software chain to create business value, so I started to reflect on that.
I got more and more inputs in my experience of silos between “Dev versus Ops”, “Product versus Dev”, “Dev versus QA”, etc. These divisions are necessary, but the interactions and systems in place are leading to all sorts of inefficiencies. I remember Nathalie Balla, co-CEO of La Redoute that led the transformation in a shadow executive committee I was participating in. One of her main frustrations and priorities was precisely to remove silos in the organization.
I also had a personal frustration in the communities I was involved in, feeling these silos. I understand that a specialized community brings focus and connects easier with the ecosystem; a “test automation” community is clearly focused for test automation, QA and test engineers. It’s great that we have this community mindset in the software industry to cope with the evolution, but I wanted to connect the dots transversally.
Based on my professional experience at Redoute, in test automation with Cerberus, with communities sharing and conferences, I reflected on the terms, leading to “Quality Engineering” with more than “only Quality” and more than “only Engineering”. I wanted to create a community with a name I would like to display and share with people. After many brainstorms and sharing, I ended up defining the Quality Engineering community as the QE Unit.
I love challenges, improving things for the future and having constant fights. Working on making Quality Engineering part of the ecosystem is something I enjoy. It’s the first reason for having the mantra “Quality Engineering or Die Trying”. It’s not like creating a community on “QA” or “DevOps” where it is easier to connect in the ecosystem. I started with different types of content like articles, meetups in round-table, podcasts and iterating on the website. I started mainly with content to share my views on the software industry and improvements in transversality, to evaluate the connection it could have in the community.
By the second half of 2021, I wanted to have more focus on a specific type of content, and better spread the word of Quality Engineering. Not being satisfied by any definition of Quality Engineering, I started to work on a whitepaper to effectively define my definition. I shared the first versions with people that were early joiners of the community, which challenged me a lot. By sharing, I started a very productive collaboration with a peer and friend of mine I truly appreciate and recognize professionally, Rémi Dewitte.
Our joint collaboration resulted in “On Defining Quality Engineering”.
I started with 100+ pages but ended with around 50 of core content. We got to a lean approach on the key messages of “Defining, Shaping & Implementing Quality Engineering”, setting up the key foundations of our Quality Engineering definition. We are now incrementally adding practices to the MAMOS framework on the site, with more content and sharing to come.
What’s your perspective of the software industry?
My summary is “Quality Engineering or Die Trying”.
Companies have an enormous pressure for digital transformation in an highly competitive and innovative ecosystem; they can only survive capturing economies of speed and scale, accelerate cycles of experimentation, streamlining software delivery, connecting better in the ecosystem with integrations and better connectivity. That’s a second reason for the mantran “Quality Engineering or Die Trying”. Too many companies still have the approach of doing things “as fast as possible”, completely putting aside necessary requirements to effectively transform and sustain changes in organizations.
We are still suffering a lot from vertical silos of expertise, like “QA ” or “Craftsmanship” or “Product Management”. While it is needed to have such focus, it lacks a transversal picture to know what to do on the end-to-end lifecycle. When we look at the entire chain, we can work on the main limiting factors, focusing on a few transformative priorities instead of addressing 10 projects per silo, when only 2 of them are necessary.
The second issue is the lack of systemic approaches in transforming organizations and systems. A lot of companies adopt for example Agile models, focusing exclusively on core methods and organizational charts, but fail to address the interactions between the teams, lack the architectural vision to reach, or to build a Quality Engineering culture. Acting partially can deliver some results on the surface, but then reality comes back to the old habits. It’s like reorganizations that change organizational charts and titles, but nothing to the day-to-day, and less to the business value delivered.
In strengths, we have progressed in the technological maturity of Cloud adoption, improvement of user experiences, connectivity between engineering components. These advances in lower layers of technology liberate more time for business alignment, innovation and deploy more business-driven technology. It’s the example with applications now giving more easily business metrics, but we still have a long way to go to improve the interoperability and connectivity. Our goal is to accelerate business with technology that enables us to test value-hypothesis with low commitment.
We can deploy CI/CD pipelines with native integrations for building the application and have unit tests executed. But when you need to add progressive deployment, multi-cloud compatibility, alerting etc things get clearly more complicated, increasing the commitment. One focus with Cerberus Testing has been to ease the deployment of test automation for different use-cases leveraging native integrations across the software lifecycle.
We have to deliver much more to accelerate companies’ transformation and ease the access to Quality at Speed software. There are many opportunities to capture leveraging data, by correctly generating, capturing and using it to provide insights and intelligent services. But this is only possible with the right foundations, that have nothing fancy or very disruptive, like standard components, data refresh, or test management. Then we get back to having a proper interconnectivity between solutions. We can have the best product available, if it can’t connect easily to other platforms it will fail to provide sufficient value.
This need for collaboration across silos, inside and outside organizations, is one main reason why I believe in industry collaboration, open-source and communities to accelerate the maturity of our industry. If we are able to share on similar problems we are having in different contexts and different maturity, we can find common solutions that are challenged more broadly in the ecosystem and provide much more value as a whole.
There is also the people, or more precisely, the Skills aspect of MAMOS. We have a failing system of training for digital transformation and technology. The market is very tense creating high turnover and other inefficiencies; schools are still training people where there are no jobs, our culture lacks a continuous training aspect that is also not supported. Some companies have developed offerings to train people and convert into digital and technology jobs, but it’s not enough if we fail to address the systemic issues.
Surfing on that wave, low-code, no-code and data science are being sold are solving the combined issues of technology and people, but I don’t believe so yet. There are a lot of foundations to set with people before reaching that, and low-code can help on narrowed verticals but cannot scale an entire business – and there still needed people with a broader vision than a very specific focus on a specific set of tasks.
Where does your inspiration come from?
I have a simple question for what I’m sharing: “Does it work?”.
This led me to share practices I have doing and my day-to-day, that can be based on existing knowledge, adapted or crafted. And that also comes from learning and sharing within the community with multiple interactions. I read 1 or 2 books per week, in mixed themes I pick on a quarterly basis that range between business, technology, leadership, inspiration, history. I complement that with a daily session of 30 minutes to learn about something I am not comfortable with, usually reading articles or watching content. I complement these inputs with regular calls with my networks, participation to the meetup, conferences, and reading also research and academic publication. I believe there is always something to learn, especially out of the comfort zone, cross areas, but also in things we already know.
More concretely in my day-to-day, I have a lot of meetings to review outputs, give feedback and my core contribution is to enable the technology transformation collaborating transversally, aligning business to IT, developing the teams and structuring them to enable changes. I use for example the One-Slide Problem Summary regularly for assessing projects or in the architecture reviews, where Minimum Viable Moves are also part of the regular method used. I systematically rely on OKRs, Kanban, DoD and DoR as fundamental methodologies to efficiently deliver results. Outside of business hours and community sharing, this is where I work focused on content creation, consolidating sources, and pushing to develop the community.
What is the future for QE Unit?
My vision is to make the QE Unit the number #1 worldwide community of Quality Engineering, regrouping people from various verticals and expertises in the software industry, with a passion for Quality at Speed software.
The community is not limited to “QA engineers” or “QA lead”, it is open to but also to engineers, architects, tech lead, engineering managers, CTOs, CIOs, CDOs; there’s no mention of titles or positions to join, who want to learn, share and contribute can join and also find a mentor.
We made a talk with Jean-François Fresi at the JFTL for “the reinvention of our quality jobs for Quality Engineering” sharing on the Quality Engineering imperative and roles evolution; that talk could be applied to the other jobs like Software Engineering or DevOps.
The community is international and sharing in English, so anyone familiar with that language can join; one dedicated channel per language is also available. It is a completely free community and I am working at developing a support model and sponsoring on the website and weekly update to accelerate and improve content creation.
I tried a variety of contents with podcasts, videos, articles etc; I now focus more on written content as I find it easier to share messages and convictions (yes I’m an introvert :)) plus a Slack for sharing messages with the community.
What are you working on at the moment?
I am focused on the transformation and team drive at La Redoute, also pushing to share and communicate outside of the company. After business hours, I keep a focus on personal development and learning as a personal driver. There is also the community involvement, being in participation, animation or contribution to the QE Unit or other communities I am involved in. In terms of companies, I am also working at developing the reach of Cerberus Testing, atale.io, and the Test Automation Camp to contribute as I can to the skills needed with test automation courses.
In July 2022, I finished the fundamental referential of methodologies for the Methods domain of MAMOS. It contains around 25 methods shared on the website sharing what it is, their contribution to Quality at Speed, and implementation guides. They are all consolidated on the page “Guides”. In next steps, I will work at improving the website design of the framework elements, and am working on a book with exclusive content in regards to the methodologies.
I am also pushing to grow the community on the mailing list and on slack to be more people and have much more sharing. The goal is to create collaboration and learning opportunities between its members, as well as mentoring or other ideas we could find valuable for the community.
I have a continuous pipeline of events I am organizing or talking at, and I have a personal objective to be able to speak of Quality Engineering as keynotes. That’s the next challenge I want to make a reality soon, at a software engineering conference.
The next content I will be publishing will be a bit more on architecture, management and organization, plus convictions I have on specific topics ; I am also collaborating to have external contributors on the site, so feel free to reach out and I can help on the writing part.
Then, I am working on making back Quality Engineering dedicated events, ideally in 2022, sharing the experience from people in the community and to make clear the Quality Engineering is the future and we should be all working on for our organizations.
Closing words by Antoine
Thanks a lot Olivier for this proposition of reversed interview, that was a good idea and I know we have common sharings and projects coming up on the theme of Quality Engineering. I would be very happy to count in for the Quality Engineering events, I believe we share the same deep convictions 🙂
I also want to thank all the people contributing to the QE Unit, yourself, Rémi Dewitte, Jean-François Fresi, Johann Gaggero, Emna Ayadi, Benjamin Butel, Iman Benlekehal, Fred Vigna, Vincent Dauce, Farah Chabchoub, Bas Vegter, and many others that could follow in this list.
Waiting for more Quality at Speed folks to join the QE Unit.