Raphaël Manfredi

• Conducted an initial capability assessment to identify the organization's DevOps maturity, followed by a comprehensive 3-year roadmap to transition to the final DevSecOps stage. • Facilitated workshops and presentations to instill DevOps principles and shared vocabulary across teams. • Challenged established practices, some of which were preventing the emergence of a DevOps culture.

• Championed Agile and DevOps culture across multiple large programs. • Coached decision-makers, including Product Owners and Program Managers, to enhance team performance and decision-making. • Developed predictive modeling to assist in realistic Program Increment scoping, enabling better priority adjustments and increased team morale.

– Re-architecture of a micro-services ecosystem in order to diminish coupling, become more RESTful, improve usage of HTTP for content and version negotiation. – Improve the ability to do Continuous Deployment. – Performance measurement and data analysis in R. – Leading rewrite of heavily loaded services to improve their performance by a factor of 10 and enable infrastructure cost savings.

– In a large French company, working for the DevOps Transformation Manager, providing strategic and tactical advice on work streams pertaining to the DevOps cultural transformation. – Coaching and mentoring in projects (developing complex Enterprise-wide information systems) to evangelize DevOps practices and culture, and optionally work on Agile practices to reinforce the natural symbiosis between DevOps and Agile: maturity assessments, workshops to establish improvement plans, explicit the lessons learned, etc.. – Worked on establishing a cross-functional technical team to cut down organizational silos, move away from ticket-based internal work processes and demonstrate that DevOps Ways and Practices can also equally apply to non-development activities. It also helped disseminate a new culture of collaboration and sharing in a population of non-developers and managers. – Conducted 2-hour sessions of DevOps presentations, face-to-face in small groups, to various population (engineers, project managers, management), in a way that was interactive to make abstract concepts easier to grasp and foster discussions between participants.

– In charge of all the R&D department with the objective to rebuild some R&D capability. – Unfortunately, when I arrived the objective was more to outsource the remaining R&D than to rebuild it, so our collaboration came to a quick end!

– Hewlett Packard Enterprise split again and I was in the part of the company that worked for the Enterprise Services business, the one that got spun out and merged with CSC to form DXC.technology. Therefore I was moved to the new company automatically. – In the process, I switched jobs and focused on DevOps Enablement within DXC, focusing on managing a team of DevOps Coaches, establishing a company-wide Release Pipeline for our service platform, along with the necessary telemetry to control and report on how well our internal automation improves or degrades performance.

– Managing development of the global IT Asset Management service, both for internal HP and for trade customers, in a cross-cultural environment. – Challenges come from unifying the services to promote code reuse and standardization, on a global scale (the Asset Management service being one component of a larger IT Outsourcing service offering). – Around 2014, I initiated a first cultural shift by moving my organization towards Agile: had people train on SCRUM, then we derived our own way of managing the flow of requirements from the business and the delivery of packages to Ops, using a mix of iterative, XP, Scrum and Kanban principles. – Starting 2015, I initiated a second cultural shift by driving the flow of work within the organization towards DevOps -- accelerating our velocity, altering wasteful processes between business, R&D and Ops, changing responsibilities to be more horizontal and incorporate Operations, the R&D and the business into one large team.

– Lead the development of an open-source Gnutella client for Unix (see http://gtk-gnutella.sf.net). Working with 3 other developers (not really employees) with IRC coordination we never met in real life. – Also participated to the evolution of the open Gnutella protocol by coming up with many proposals and extensions to the protocol and coordinating the specifications so that other clients could inter-operatate. – Nowadays (since about 2009) my work there merely involves low-level system programming since the Gnutella protocol is no longer evolving. – I developed several memory allocators to be able to run the program for months without fragmenting memory, even replacing the native malloc() with my own allocator, I developed a thread layer with features to assist in deadlock detection and debugging, offering inter-thread communication and inter-thread function calls. – I also developed an original POSIX emulation layer on Windows to port the program to that platform whilst minimizing the changes to the source code and yet get a native program, not a Cygwin one. – For Windows, I also implemented my own backtrace() routine, necessary to be able to trace a stack across Windows routines and gcc-compiled ones -- they use a different calling conventions -- involving on-the-fly disassembly of the routine startup code, to be able to compute the amount of stack used by that routine based on the way it manipulates the SP register.

– Managing the teams within the EMEA region responsible for implementing and providing technical support of the IT Asset Management service sold to trade customers. – Accountable for all technical matters in the service implementation and for diminishing the overall cost of the solution whilst retaining a high quality of service, so as to make it very competitive on the market.

– Working in the IT Asset Management area, in a team responsible for the whole solution: starting from pre-sales investigations to actual delivery of the service. – Responsible for pre-sales deal pursuit (understanding customers' needs, identifying the risks, costing) and also for following the actual transformation activities of customers for which we are starting a deal (project management activities).

– Lead architecture and development of a multi-tier integration portal with a web interface offering a unique interface to customers, giving them access indirectly to HP's support infrastructure. – The customers of this service are big accounts which chose to outsource the management of their IT to HP. – I also conducted system performances analysis (CPU load, cache policy impacts, bandwidth and throughput analysis, memory footprint, etc...) as well as pro-active monitoring and support 24x7 of the production systems.

– The Perl Storable module is a very fast persistance mechanism for Perl, written in C for speed and sacrifying portability to efficiency by directly attacking the low-level data structures of the Perl interpreter instead of going through the public core API. – Since June 2001, this module has been part of the mainstream and it is now maintained directly by the Perl authors.

– Worked in the software engineering department. – Assessor for the CMM level 2 evaluation of the organization. – Architected and lead development of a web-based time reporting tool for software developpers and project leaders.

– Teaching Operating Systems to third year students, 30 hours per year.

– Worked in the team building ATM drivers for HP-UX. – Lead the internal quality of the drivers by providing a simulating platform for drivers, including a user-mode port of the HP-UX kernel (like what is done today for user-mode linux). – Also worked on the re-architecture of a real-time database replication module via TCP/IP for a telecom application (Intelligent Networks), to allow 400 times more throughput through the same link.

– Worked as a sub-contractor for HP, to optimize the performances of an existing telecom software, managing GSM Home Location Register (HLR) platforms.

– Worked in the DEC/OSF1 Unix kernel to port it on the ACRI architecture, with a few specific developments (interrupt layer, libc interface for setjump/longjump, signal trampolines, file system adaptations)

– Participated to the writing of the ISE Eiffel 3 compiler. I focused mainly on the low-level aspects: runtime (written in C, responsible for memory management, garbage collection, exception handling and code reflection), code generation (written in Eiffel). – Eiffel is an object-oriented language invented by Bertrand Meyer. ISE is the company founded by Bertrand Meyer to promote the Eiffel language and the related Object-Oriented concepts. The company was renamed "Eiffel Software" in 2001 or so.