Sr. AI Native Engineer
- Guadalajara, MX Guadalajara, Mexico (Hybrid) Long term contract Client-Virtusa/Google
• We are seeking high-performing AI-Native Engineers across Senior and Intermediate levels to
• build, operate, maintain, and continuously improve enterprise Generative AI, Agentic AI,
• Multi-Agent systems, and data-centric AI solutions. These are core engineering and production
• operations roles—not research-only or prompt-engineering-only positions.
• In this role, engineers will own and drive the complete lifecycle of enterprise AI implementations
• across Development → Integration → Testing → Deployment → Observability →
• Production Support → Incident Resolution → Optimization → Data Migration →
• Continuous Evaluation. You will be directly responsible for supporting production-grade
• Agentic AI solutions leveraging Gemini, Google Agent Development Kit (ADK), Vertex AI,
• BigQuery, RAG, Vector Databases, GCP, AI observability frameworks, data pipelines, and
• Google-native production infrastructure. 2. Senior vs. Intermediate Level Expectations Scope & Autonomy Architects complex multi-agent interactions, leads incident resolution, and owns high-availability architecture and end-to-end continuous evaluation frameworks. Leadership & Ownership Drives post-incident Root Cause Analysis (RCA), sets CI/CD & safety deployment standards, and leads migration architecture. Problem Solving Solves complex probabilistic behavior failures, cross-system latency/quota throttling, and data/model drift issues. 3. Core Responsibilities 1. Agentic AI Engineering
• Develop, deploy, and maintain robust Multi-Agent systems and application workflows.
• Architect and construct supervisor/sub-agent orchestration patterns.
• Implement agent tools, custom extensions, and API integrations.
• Engineer deterministic execution paths surrounding probabilistic LLM reasoning to ensure
• system reliability.
• Design and implement fail-safe mechanisms, retries, and fallback recovery paths.
• Maintain and version agent prompts, system configurations, and schema tool definitions.
• Troubleshoot complex agent execution failures and execution loops.
• Support, optimize, and maintain RAG (Retrieval-Augmented Generation) and vector-search
• pipelines.




• Implement Human-in-the-Loop (HITL) workflows to enable seamless human intervention and
• verification.
• Support continuous learning frameworks and automated feedback loops 2. Production Maintenance & Operations
• Engineers possess direct, end-to-end responsibility for production system health and 24x7
• resilience. Activities include:
• Conducting daily operational health checks and proactive production monitoring.
• Executing rapid incident triage, log analysis, and Root Cause Analysis (RCA).
• Performing prompt adjustments, configuration changes, bug fixing, and pipeline reruns.
• Investigate and resolve API execution failures, quota limits, throttling, and latency spikes.
• Managing production deployments, shadow launches, and rollback operations.
• Conduct post-incident analysis and implement preventive operational controls. 3. AI Observability & Telemetry
• Build and manage comprehensive observability frameworks for Agentic AI applications,
• monitoring key telemetry such as:
• Agent Operations: Invocation count, latency, task-completion rates, failure rates, fallback
• rates, routing accuracy, and insight acceptance.
• LLM & Tool Performance: LLM call latency, tool/API error rates, token consumption trends,
• and quota consumption.
• Data & Model Quality: Data quality, schema drift, model/extraction drift, prompt regressions,
• and HITL feedback scores.
• Tooling

Experience:
Hands-on experience with Monarch, PLX, Cloud Monitoring,
• OpenTelemetry, or equivalent enterprise observability systems. 4. Data Migration & Data Engineering
• Participate actively in data migration and modernization initiatives, ensuring data integrity for
• downstream AI behavior:
• Executing legacy data migrations, BigQuery migrations, and schema modernizations.
• Developing ETL/ELT data pipelines, data transformation, validation, and reconciliation
• scripts.



Performing source-to-target quality validation and semantic-layer migrations (LookML, data
• models).
• Managing Vector DB and RAG knowledge-base data migrations and metadata
• synchronizations.
• Implementing enterprise data retention, purging, and archival policies. 5. RAG & Vector Database Operations
• Maintain ingestion and embedding pipelines; execute re-indexing upon source data mutation.
• Validate vector ingestion accuracy, metadata completeness, and freshness of knowledge
• bases.
• Investigate retrieval failures, inaccurate context insertion, and ranking anomalies.
• Support automated RAG regression testing and maintain synchronization between source
• systems, BigQuery, and Vector DBs. 6. Continuous Evaluation & QA
• Establish and maintain Golden Datasets for continuous benchmark testing.
• Execute Shadow Mode testing, Side-by-Side (SxS) output validations, and automated
• prompt evaluations.
• Implement hallucination detection, output drift monitoring, and HITL acceptance analysis.
• Systematically isolate issue origins across: Data → Prompt → Model → Agent Logic → Tool→ Retrieval → Infrastructure → Integration. 7. High Availability, Resilience & Security
• Implement idempotency, timeouts, retry budgets, circuit breakers, dead-letter queues (DLQ),
• and asynchronous processing.
• Manage capacity, autoscaling, and graceful degradation to maintain highly available 24x7
• services.
• Enforce security controls including PII masking/redaction, Identity and Access Management
• (IAM), RBAC/ACL, secret management, audit logging, and automated compliance retention. 4. Technical Qualifications & Stack
• AI & Agentic Frameworks: Gemini, Google ADK (Agent Development Kit), Vertex AI,
• LangChain/LlamaIndex (or equivalent multi-agent frameworks).
• Data & Cloud Infrastructure: GCP (Google Cloud Platform), BigQuery, LookML, Vector
• Databases (Vertex AI Vector Search, Pinecone, Chroma, etc.), ETL/ELT pipelines.
• Observability & Monitoring: Monarch, PLX, Cloud Monitoring, Cloud Logging, Prometheus,
• Grafana, Datadog, or equivalent.
• Software Engineering & CI/CD: Python, REST/gRPC APIs, Docker, Git, CI/CD pipelines,
• automated testing, version control for prompts/configs.

📌 Sr. AI Native Engineer Guadalajara, MX
🏢 xchange
📍 Guadalajara