Core Competencies
DI-DS delivers production artificial intelligence across five core competency areas: surveillance and video analytics, unmanned systems and edge AI, custom AI application development and modeling, agent orchestration, and AI governance and assurance. Each competency is supported by systems the company has developed and currently operates, and by the principal's prior delivery in defense, intelligence and regulated commercial environments.
Surveillance and Video Analytics
Person and object detection, multi-object tracking and re-identification, pose estimation and action recognition. Multi-camera synchronization and homography-based spatial alignment across differing viewpoints. GPU-accelerated distributed inference on production systems processing millions of video events daily. Capabilities applicable to ISR, force protection, perimeter security and physical security monitoring.
Unmanned Systems and Edge AI
AI inference deployed to NVIDIA Jetson aboard unmanned aerial and ground platforms. ROS2 integration for autonomous navigation and real-time perception. ONNX export, structured pruning and graph optimization for resource-constrained embedded GPU. RF signal classification and automatic modulation recognition developed for Department of Defense drone systems.
Custom AI Application Development and Modeling
Complete production AI applications, not prototypes. Full-stack systems combining machine learning inference, backend services and user interfaces, deployed on AWS and Azure with automated CI/CD via Azure DevOps and GitHub Actions. Purpose-built models where commercial off-the-shelf does not fit: LLM fine-tuning on curated domain corpora, document intelligence combining OCR and NLP, hyperspectral imaging analysis.
Agent Orchestration
Agentic AI architectures: task decomposition, multi-agent coordination, tool use and function calling, human-in-the-loop validation, and deterministic control layers that keep language models out of safety-critical decision paths. Retrieval-augmented generation with citation validation and structured output enforcement. Azure AI Foundry implementation.
AI Governance and Assurance
Establishes and leads an enterprise AI Center of Excellence setting AI strategy, standards, model risk documentation and review processes in an FDA-regulated medical device environment. Governance experience spanning FDA, HIPAA and SOC 2. LLM observability and traceability, independent model evaluation and adversarial testing of agentic systems.
Production systems, internally funded
DI-DS has designed, built and currently operates production AI systems with its own investment. The figures below were measured on Goatie, the company's computer vision and records platform, as it runs today.
- On-device object detection and animal re-identification at 6.5 ms per frame, AP50 0.88, on a hybrid edge and cloud architecture.
- Explainable inference in production: every model output carries its reasoning and a confidence level, and the system is positioned as decision support.
NAICS and PSC codes
- 541511
- Custom Computer Programming Services
- 541512
- Computer Systems Design Services
- 541519
- Other Computer Related Services
- 541715
- R&D Physical, Engineering and Life Sciences
- 541690
- Other Scientific and Technical Consulting
- 541990
- Other Professional, Scientific and Technical Services
- 518210
- Computing Infrastructure and Data Processing
- PSC
- B506 · DA01 · DA10 · DF01 · R425 · R499
Core technologies
- Modeling
- Python, PyTorch, TensorFlow, ONNX Runtime, OpenCV, MediaPipe, Detectron2
- Edge
- NVIDIA Jetson, CUDA
- AWS
- SageMaker, Bedrock, Textract
- Azure
- Azure DevOps, Azure AI Foundry
- Platform
- MLflow, Docker, Terraform
- Data
- Milvus, pgvector, PostgreSQL, Kafka, Airflow
- Personnel
- No active personnel or facility clearance. Principal previously held Top Secret (USAF). Clearance-eligible with sponsorship.
Differentiators
- Surveillance and unmanned systems specialization with production depth: person and item tracking with action recognition delivered in production at Fortune 10 retail scale, plus Department of Defense work on Vision Transformer RF signal classification for unmanned aerial systems (UAS).
- Prior United States Air Force service under a Top Secret clearance, with three years as a defense contractor in Kuwait and Afghanistan.
- Published deep learning research: curriculum learning for automatic modulation classification (ProQuest, 2023) and IEEE World AI IoT Congress (2022).
- Production machine learning delivered inside HIPAA, SOC 2 and FDA-regulated environments, including standing up the AI governance and model risk protocols themselves.
- Multi-cloud delivery across both AWS and Azure rather than a single-vendor practice. Principal holds an M.S. in Computer Science.
Capability statement
The DI-DS one-page capability statement summarizes core competencies, past performance, differentiators and company data for contracting officers and teaming partners.