> Markdown version of [/jobs/ext/3931674-ai-assisted-cyber-analysis-specialist](https://www.wearedevelopers.com/jobs/ext/3931674-ai-assisted-cyber-analysis-specialist). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# AI-Assisted Cyber Analysis Specialist

- **Company:** CACI International Inc.
- **Location:** Reston, VA, United States
- **Experience:** Expert
- **Salary:** $90,300.0 - $189,600.0
- **Contract:** Permanent contract
- **Skills:** Application Programming Interfaces (APIs), Artificial Intelligence, Amazon Web Services, CompTIA Security+, Continuous Integration, Data as a Services, Information Engineering, Information Leak Prevention, Data Stores, Fuzz Testing, Python (Programming Language), Machine Learning, Search Technologies, Data Processing, Retrieval-Augmented Generation, Large Language Models, Software Security, Model Validation, Kubernetes, Data Management, Azure AI, Splunk, Static Application Security Testing, Vulnerability Analysis, Dynamic Application Security Testing
- **Published:** October 26, 2026
- **Apply:** https://www.techcareers.com/job.asp?id=3467377234&tx=CT3028TYV&pt=1&aff=0B19D771-A501-4A5E-8338-2A822B784D54&utm_source=Job%20Feed&utm_medium=textkernel&utm_campaign=DE&utm_term=0B19D771-A501-4A5E-8338-2A822B784D54

## About the Role

* U.S. citizenship is required.
 * The selected candidate must meet eligibility requirements for access to sensitive information and be able to obtain a Public Trust fitness determination (High Risk).
 * Ability to work in customer-provided remote environments, use customer-approved tools, and comply with rules of engagement, data-handling requirements, evidence controls, deconfliction procedures, and stop-work criteria.
 * Five or more years of combined cybersecurity analysis, security engineering, software/data engineering, or applied AI/ML experience, including production automation used by analysts or operators.
 * Strong Python, API, data-processing, testing, and source-control skills; experience handling structured and unstructured technical artifacts at scale.
 * Practical knowledge of vulnerability assessment, attack paths, source-code or configuration analysis, malware/artifact triage, or security-tool outputs sufficient to validate analytical results.
 * Experience building and evaluating AI/ML or LLM-assisted workflows with explicit human review, grounding/provenance, structured outputs, and measurable quality controls.
 * Understanding of AI security risks including data leakage, prompt injection, untrusted content, model hallucination, authorization, auditability, and secure deployment.
 * Ability to communicate limitations and uncertainty clearly and to produce maintainable technical documentation.

Desired:

 * Experience with agentic workflow frameworks, retrieval-augmented generation, vector search, model evaluation/observability, or secure model gateways in approved enterprise environments.
 * Experience with Semgrep, CodeQL, YARA, Sigma, SAST/DAST, SBOM, malware-analysis, vulnerability-management, or exposure-management data.
 * Familiarity with MITRE ATLAS, NIST AI RMF, secure AI development, adversarial testing, and AI red-team practices.
 * AWS/Azure AI and data services, container/Kubernetes, CI/CD, data-lake, Splunk/Elastic, or graph-analytics experience.
 * Relevant cloud, data, AI/ML, or cybersecurity certifications.

## Description

* Design human-in-the-loop AI-assisted workflows to classify, enrich, correlate, summarize, and prioritize large volumes of code, configurations, scan results, findings, evidence, and other cyber artifacts.
 * Build Python services, tools, APIs, prompts, schemas, retrieval pipelines, and evaluation harnesses using customer-approved model endpoints and data stores.
 * Integrate deterministic security tools and rules-such as SAST/DAST, Semgrep, YARA, Sigma, SBOM, vulnerability, and configuration results-with AI-assisted triage rather than relying on model output alone.
 * Measure precision, recall, false-positive and false-negative rates, latency, cost, drift, and analyst time saved; document acceptance thresholds and rollback paths.
 * Protect sensitive data by enforcing access controls, approved model boundaries, provenance, audit logs, redaction, prompt-injection defenses, and safe handling of untrusted code and artifacts.
 * Present structured findings and uncertainty to security analysts, support manual validation, and capture analyst feedback for continuous improvement.
 * Maintain technical documentation, test datasets, model/workflow versions, and reproducible release processes.