Title: MLOps Engineer
Duration: Contract
Location:- Sacramento, CA
Role summary
The MLOps Engineer makes AI applications reliable in production and keeps them healthy afterward. This is the role that closes the gap most AI projects die in: Day-2 operations. It is distinct from infrastructure-level platform operations (GPU provisioning, cluster tuning), which partners may own. This role focuses on application-level operations: deployment automation, evaluation pipelines, and monitoring for the failure modes specific to LLM systems, across whichever platform the application runs on.
Key responsibilities
- Build and operate CI/CD pipelines for AI applications, including containerized deployment.
- Stand up evaluation pipelines using RAG-specific metrics (faithfulness, context precision and recall, answer relevance via RAGAS or equivalent).
- Monitor the failure modes that matter for LLM systems: hallucination rate, knowledge drift (content changed, index stale), retrieval-quality degradation, prompt-injection attempts, and latency trends.
- Define and track production metrics, report against baselines (use percentiles, not just averages), and maintain audit-ready logging that feeds governance reporting.
- Operate model and content refresh: re-indexing, embedding refreshes, model-version upgrades, and guardrail tuning as the environment evolves.
- Own incident response for AI failures and contribute to the Day-2 managed-services capability.
- Distinguish application-level issues from infrastructure-level ones and maintain a clean interface with infrastructure owners.
Required qualifications
- Strong DevOps foundation: containers (Docker), CI/CD, infrastructure-as-code, and observability tooling.
- Experience monitoring ML or AI systems in production, not only traditional application monitoring.
- Familiarity with experiment tracking and evaluation tooling (for example, MLflow, LangSmith, RAGAS).
- Understanding of LLM-specific failure modes and why they differ from classical ML drift.
- Comfortable with Python and scripting for automation and custom evaluation logic.
Multi-platform requirement
Must deploy, monitor, and operate across all three:
- Azure AI: Azure ML pipelines, Azure Monitor, deployment to Azure-hosted endpoints.
- AWS AI: SageMaker pipelines, CloudWatch, Lambda and container-based deployment.
- On-prem NVIDIA AI factory: operating applications served via Triton and NIM, working with Run:ai scheduling at the application boundary, and monitoring on-prem deployments where cloud-native observability tooling is not available.
About Logisoft:
We represent Logisoft Technologies, Inc. with pride; We present ourselves as a premiere Technology, Consulting, Product Development, and Software Services Company.
Our Head Office is located in South Plainfield, NJ, and Our Offices location is in Hyderabad, INDIA & Accra, GHANA.
We are Microsoft Official Partners - A Microsoft Certified Partner that helps customers with a range of IT projects and specific IT solutions.
We started our journey in the year 2008. In the past 17+ years, we have acquired the trust of many IT market leaders.
LOGISOFT Technologies is committed to a policy of Equal Employment Opportunity and will not discriminate against an applicant or employee on the basis of age, sex, sexual orientation, race, color, creed, national origin, ancestry, disability, marital status, or any other legally protected basis under federal, state or local law.
Interested candidates can apply directly to this posting or they can reach
Never miss an opportunity! Create an alert based on the job you applied for.
