Steelcut Software, Inc. offers services for advanced software system design, development, integration, and maintenance.

Prototype Reengineering, Stabilization, and Hardening

Basic software prototypes can save time and resources, but are not usually production-ready without further planning and adaptation. In focusing on speedy proof of concept, developers may skip over key considerations such as testing, security, performance, scalability, and maintainability. Steelcut Software excels at filling in these gaps. Our streamlined processes, based on years of experience and dedication to continuous improvement, reduce the cost of adapting prototypes for production.

Steelcut’s prototype-to-production efforts include:

  • Planning for incremental adaptation of capabilities and delivery to production
  • Establishing clear, consistent data migration and issue resolution procedures
  • Incorporating automated unit, integration, and functional tests
  • Automating build and delivery tasks
  • Tuning underlying tools and servers for production deployment

Steelcut Software’s dedication to service and maintaining relationships ensures a focus on going the distance with our partners. Essential services include:

  • Monitoring system performance
  • Tracking and rapidly addressing security concerns
  • Identifying opportunities to help you meet your new challenges

Production-Focused Data Science and Engineering

Production-grade Machine Learning models impose more requirements than experimental and evaluation models. Steelcut Software performs focused data science and engineering while maintaining real-life production goals.

Our Artificial Intelligence solutions:

  • Use the smallest, fastest, least complex algorithms that produce good results
  • Ensure models are scalable to production requirements
  • Institute model versioning
  • Establish model performance metrics
  • Automate selected data engineering activities
  • Evaluate production performance
  • Plan for data collection and periodic retraining
  • Protect sensitive data used in models

DevOps, DevSecOps, and MLOps Planning and Implementation

Automation is critical to minimize regression issues and ensure reliable, repeatable delivery. Automation touches all areas of the production software landscape, including security, machine learning experiments and model creation, and software development, delivery, and operations.

Steelcut Software has 20 years' experience in planning and implementing automation of the following processes:

  • Unit, integration, and functional testing with coverage analysis
  • Software bundling and versioning
  • Data engineering
  • Machine Learning model performance analysis, experiment tracking, and versioning
  • Code analysis
  • Software deployment
  • System monitoring and notification

Cybersecurity Requirement Integration

Steelcut Software has extensive knowledge of the cybersecurity requirements that DoD contractors must satisfy for the DFAR. Steelcut Software is compliant with NIST Special Publication 800-171: Protecting Controlled Unclassified Information in Nonfederal Systems and Organizations.

Steelcut Software also has experience creating, deploying, and maintaining systems that operate in environments meeting DoD Cyber Exchange STIGs and SP 800-171 standards. In such systems, Steelcut takes an active role, even identifying and adopting best practices beyond base requirements, where practical. Examples include:

  • Monitoring and rapidly responding to reported cybersecurity vulnerabilities (CVEs)
  • Considering and monitoring a system's software supply chain
  • Adopting best practices in web application development outlined by the Open Web Application Security Project (OWASP)

To reduce the cost of maintaining a defensive security posture, Steelcut Software applies the most advanced tools available. Steelcut automates processes where possible, such as:

  • Using OpenSCAP to ensure configuration compliance
  • Automating STIG configuration baseline creation and vulnerability scanning
  • Leveraging the the ELK stack — Elasticsearch, Logstash, and Kibana — for system monitoring

About

Ray Budd - CEO

Ray has been developing advanced software as a government contractor for more than 20 years. His lengthy experience on such programs has given him a solid understanding of the DoD's contracting requirements and processes.

Ray has comprehensive knowledge of transforming systems from prototype to production, as well as extensive experience in updating production systems with minimal risk and impact to end-users. He has a background in production-oriented data science and has led the development of multiple successful web applications backed by advanced ML models.

Ray is the co-author of two books, and has had multiple publications in refereed journals and conferences.