Solution Architect

Roles & Responsibilities

Role Summary

The AI Solution Architect will lead the design and delivery of secure scalable AI solutions spanning Generative AI agentic AI machine learning data and cloud. The role will architect governed AI agents built on robust agent loops and harnesses with guardrails human oversight and observability that execute multi-step business tasks reliably. Working with business engineering and governance teams the architect will take AI use cases from concept to production.

Key Responsibilities

Define end-to-end AI solution architectures aligned with business outcomes enterprise standards and technology strategy.

Assess business use cases and recommend suitable approaches using Generative AI machine learning natural language processing computer vision predictive analytics or intelligent automation.

Design architectures for large language model applications including prompt orchestration retrieval-augmented generation agents vector databases model gateways and enterprise integrations.

Design agentic AI solutions built on the reasonactobserve agent loop with defined step limits exit conditions and escalation paths for reliable multi-step task execution.

Architect agent harnesses covering context engineering tool orchestration (including Model Context Protocol) state persistence checkpointing and output verification.

Embed human-in-the-loop approvals sandboxed tool execution and cost retry and resource guardrails into agentic workflows.

Establish agent evaluation tracing and monitoring practices to measure task success reliability cost and failure patterns in production.

Evaluate and select foundation models AI platforms cloud services frameworks and tools based on performance security scalability cost and compliance requirements.

Establish reusable reference architectures design patterns accelerators and guardrails for responsible AI adoption.

Design data ingestion preparation feature embedding retrieval model-serving API and application-integration layers.

Guide engineering teams on model development fine-tuning evaluation deployment monitoring and lifecycle management.

Define MLOps and LLMOps practices covering versioning automated testing deployment pipelines observability drift detection rollback and continuous improvement.

Embed security privacy explainability fairness human oversight intellectual-property protection and regulatory controls into solution designs.

Lead architecture reviews technical discovery workshops proof-of-concept initiatives and production-readiness assessments.

Create architecture diagrams solution blueprints technical proposals effort estimates implementation roadmaps and risk-mitigation plans.

Engage with customers and senior stakehold ers to communicate solution value design choices dependencies risks and trade-offs.

Mentor technical teams and contribute to AI capability development knowledge assets and engineering standards.

Technical Competencies

Generative AI: Large language models prompt engineering retrieval-augmented generation agents model evaluation fine-tuning guardrails and vector search.

Agentic AI: Agent loop design (ReAct and plannerexecutor patterns) agent harness engineering multi-agent orchestration tool calling and Model Context Protocol memory and state management human-in-the-loop controls and agent evaluation and observability (e.g. OpenTelemetry tracing).

AI/ML: Supervised and unsupervised learning deep learning natural language processing computer vision forecasting feature engineering and model validation.

Platforms and Frameworks: Azure AI Foundry/Azure Machine Learning AWS Bedrock/SageMaker Google Vertex AI OpenAI-compatible services Hugging Face LangChain LlamaIndex or similar technologies.

Data and Integration: SQL and NoSQL databases data lakes or lakehouses vector databases APIs messaging ETL/ELT metadata and data governance.

APIs and Microservices: RESTful GraphQL OpenAPI specifications API gateways and management microservices and event-driven architecture service mesh asynchronous messaging (e.g. Kafka) API security (OAuth 2.0 OIDC JWT) and exposing AI models and agents as reusable scalable services.

Engineering and Operations: Python containerization Kubernetes CI/CD MLOps LLMOps model monitoring observability and infrastructure as code.

Architecture and Governance: Cloud-native design security architecture identity and access management scalability resiliency FinOps responsible AI and regulatory compliance.

Preferred Qualifications

Experience in consulting presales solution estimation proposal development or customer-facing architecture roles.

Experience delivering AI solutions in regulated or large-scale enterprise environments.

Knowledge of industry-specific AI use cases particularly in telecommunications customer operations quality engineering network operations or business observability.

Experience establishing AI centers of excellence reusable AI platforms or enterprise AI governance frameworks.

Education

Bachelors degree