T
Torch.AI builds foundational AI infrastructure for U.S. defense operations, structured around three core products: ORCUS (data movement), NEXUS (semantic vectorization), and HALO (graph-native reasoning). The tech stack—Python, PyTorch, TensorFlow, LangChain, PostgreSQL, MongoDB, Elasticsearch on AWS—reflects a mature ML engineering organization. Hiring velocity is accelerating with 13 engineering roles and a mid-to-senior skew, while pain points center on operational constraints (intermittent connectivity, low-swap hardware, seconds-to-decision timelines) that require both sophisticated reasoning and edge-deployable inference.
Tech Stack 24 technologies
Core StackPython FastAPI Flask Java Apache NiFi scikit-learn PyTorch TensorFlow LangChain PostgreSQL MongoDB Elasticsearch AWS Selenium Cypress JUnit Postman NEXUS Spring Boot Micronaut spaCy Haystack Transformers pytest
What Torch.AI Is Building
◆Challenges
- Operational ai under real constraints
- Intermittent connectivity
- Degraded visibility
- Adversarial mis/disinformation
- Maintaining ato accreditation
- Managing vulnerabilities
- Low-swap hardware constraints
- Seconds-to-decision timelines
- Messy adversarial multi-domain data
- Secure modernization efforts
▲Active Projects
- Nexus semantic vectorization
- Ai control layer platform
- Anomaly detection
- Orcus data movement system
- Orcus data movement platform
- Predictive movement analysis
- Multi-int fusion
- Ai control layer deployment for u.s. government
- Ai control layer development
- Targeting workflows
Hiring Activity
20 roles · 8 in 30d
Department
Seniority
Company intelligence