_________________      ____
  / ____/ ____/ ___/     / / /____
 / /   / /    \__ \     / / / ___/
/ /___/ /___ ___/ /    / / / /__
\____/\____//____( )  /_/_/\___/
                 |/
brett.estrade@coastalcomputingservices.net                +1-281-595-1886 (text first)
Through Coastal Computing Service, LLC., Brett Estrade provides customized programming (specializing in Perl and "Unix" shell scripting), devops, backend systems, data processing, and operational support with a particular focus on real time ADCIRC storm surge model automation, optimization, and HPC cluster operations. Work also includes maintaining production Linux environments, integrating new compute architectures into existing clusters, developing backend and data-processing systems, and maintaining long-lived business and scientific applications. The founder of CCS, LLC is one of the original developers of the ASGS (ADCIRC Storm Guidance System) and currently serves as the lead developer and architect of this system.
Menu Rates
Specialties include:
  • Perl Development and Consulting - 24+ years of experience:
    • extend or fix legacy Perl cgi-bin
    • legacy Perl code organization, maintenance, refactoring, modularization
    • creating tools/utilities
    • state of the art Perl applications (e.g., scalable SaaS/web APIs, clients)
    • backend application development and maintenance
    • database-backed Perl systems and administrative applications
    • authentication, registration, subscription, and access-control workflows
    • beginner-advanced training
    • CPAN module maintenance, creation, testing, portability, etc
    • Perl module build, XS, compiler, and cross-platform troubleshooting
    • ETL & web scraping
    • large-scale data ingestion, cleanup, reconciliation, and transformation
    • processing and maintaining historical datasets accumulated over many years
    • automated email, reporting, and publishing pipelines
    • scientific computing and systems automation using Perl
    • Pretty much anything else (note: degenerates, purveyors of immoral things, and usurers need not contact me)

    Note: Being a Perl dev today is a lot like being the ac guy or mechanic who realizes people just want to keep their stuff working or tweak things ever so slightly in a very cost effective way. As they say; I'll fix what you've got. :-)

  • Professional ADCIRC Consulting - https://ADCIRC.live
    • Installing ADCIRC on most platforms
    • Porting ADCIRC to new platforms
    • ADCIRC automation and scripting
    • ADCIRC performance optimization, profiling, and benchmarking
    • Benchmarking and scaling studies
    • MPI decomposition and mesh partition optimization
    • ADCIRC+SWAN coupled-model performance testing and optimization
    • MPI instrumentation and analysis of computational and communication performance
    • Managing and conducting large amounts of ADCIRC "runs"
    • Creating repeatable benchmark frameworks for comparing ADCIRC configurations
    • Creating custom tools, commandline and online SaaS platforms

    Note: I am not a domain scientist or engineering consultant, but work with one of the best @ ADCIRC.live.

  • Scientific and High Performance Computing - 20+ years of:
    • OpenMP (gcc/GOMP, intel, pgi, etc)
    • MPI (OpenMPI, MPICH, Intel MPI)
    • building/porting code (C, Fortran; autoconf, make)
    • batch/queue systems, including Slurm
    • building, configuring, upgrading, and maintaining Linux HPC clusters
    • integrating new compute nodes and architectures into existing production clusters
    • Slurm controller, compute node, GPU node, and heterogeneous cluster integration
    • x86-64 and ARM/AArch64 HPC systems
    • adding ARM/GPU compute capacity to established x86-64 Slurm environments
    • GPU resource discovery, scheduling, and GRES configuration
    • GNU and Intel oneAPI compiler and MPI environments
    • shared software/toolchain environments using NFS and Lmod
    • MPI process placement, affinity, fabric configuration, and multi-node performance tuning
    • CPU, memory, storage, MPI, and application-level benchmarking
    • post-outage recovery and restoration of multi-node scientific computing environments
    • diagnosing scheduler, filesystem, software, and compute-node failures in production clusters
    • Extensive experience supporting domain scientists running the ADCIRC model on many platforms and in many configurations

    Note: See OpenMP::Simple, an approach to using OpenMP and Perl together.

  • Additionally:
    • Unix shell scripting
    • Linux server and production application maintenance
    • Slurm and Linux HPC cluster administration
    • Ansible-based system and cluster configuration
    • backend web application and API support
    • database administration, migration, and data repair
    • ETL, data warehousing, and archival-data processing
    • automated ingestion and normalization of externally sourced data
    • large historical datasets and recurring data-processing workflows
    • production email, subscription, registration, and reporting systems
    • Linux/cPanel server administration, recovery, hardening, and troubleshooting
    • web application performance and abuse/bot mitigation
    • legacy production system maintenance where replacement is impractical or unnecessary
    • Containerization: FreeBSD Jails, FreeBSD bhyve, Docker
    • Git-based development workflows, code reviews
    • Perl training,

Flexible working/payment arrangements available. Always confidential and discreet. For general inquiries about availability, please email me. References available on request from clients who have consented.

Coastal Computing Services, LLC © 2021-2026

DBA: Acutis Computing & Data Consultancy, LLC

+1-281-595-1886 (text first)