Resident Engineer, AI Security

Resident Engineer, AI Security The Resident Engineer, AI Security serves as a hands-on technical engineer embedded onsite with one of A10's largest financial services customers in Sunrise, Florida. You will sit directly with the customer's AI platform, model risk, and security engineering teams to operate and scale two critical disciplines: runtime AI firewall enforcement and adversarial model testing.

On the runtime side, you will deploy and operate inline inspection of AI traffic — prompts, responses, retrieval context, tool/MCP calls, and agent-to-agent exchanges — managing policies that determine what gets blocked, redacted, rewritten, or logged. On the testing side, you will build and run adversarial evaluation workflows that probe models, RAG pipelines, and agents before and after they reach production.

This role suits an intermediate-level engineer who bridges backend systems development, cloud infrastructure, and practical AI security.  
** Responsibilities*
* ** AI Firewall and Runtime Enforcement
*** Deploy, configure, and maintain inline AI traffic inspection across inference paths (self-hosted models, commercial APIs, RAG, and agents)
* Manage proxy topologies, TLS termination/SSLi, streaming/SSE response handling, and proxy routing across modern network stacks using tools like Envoy, NGINX, or HAProxy
* Implement and tune guardrail policies: prompt injection/jailbreak detection, PII/PCI/MNPI data leakage, grounding controls, and token rate limiting
* Enforce authorization, tool-use policies, and identity attribution across REST and gRPC endpoints to ensure every agentic action maps to a verified principal
* Monitor baseline false positive/negative rates, shadow-test policy updates, and manage inspection latency to meet production SLA budgets  
** Adversarial Model Testing and Red Teaming
*** Execute automated red team campaigns against models, vector databases (e.g., PostgreSQL with pgvector), and RAG pipelines covering injection, data extraction, poisoning, and cost-exhaustion risks
* Maintain versioned test suites and automated evaluation harnesses using Python and Go
* Integrate security testing gates directly into the customer’s MLOps, CI/CD, and dev pipelines
* Produce structured finding reports with clear reproduction steps, severity assessments, and concrete remediation steps (policy changes, system prompt updates, or infrastructure controls)  
** Customer Collaboration & Product Feedback
*** Collaborate daily with the customer's platform architects, model risk teams, and security leads on control selection and threat modeling
* Map control coverage to common industry frameworks (OWASP Top 10 for LLMs, MITRE ATLAS, NIST AI RMF)
* Feed operational feedback, detection gaps, and bug reports back to A10 Product and Engineering teams to continuously improve the core product  
** Required Qualifications**
* *
* Experience:

** 5–7 years of professional experience in software engineering, security engineering, cloud platform engineering, and technical solution delivery
* ** Programming

Languages:

** Proficiency in
** Python
* * and
** Go
* * for building production-grade integrations, microservices, micro-benchmarks, and test automation harnesses
* ** Data & Middleware:
** Strong working knowledge of
** PostgreSQL** (schema design, query optimization, indexing) and API architectures using
** REST
* * and
** gRPC**
* ** Cloud & Dev Ops:
** Hands-on experience deploying and managing workloads in at least one major cloud provider (
** AWS, GCP, or Azure**), alongside containerization (Docker, Kubernetes) and CI/CD pipelines
* ** Networking & Traffic Management:
** Solid understanding of core networking concepts, L4/L7 traffic management, reverse proxies (e.g., Envoy, NGINX, HAProxy), TLS…