- Company Name
- ConSol Partners
- Job Title
- AI Application Architect
- Job Description
-
Job title: AI Application Architect
Role Summary: Design, deploy, and govern end‑to‑end AI/ML solutions on cloud platforms, transforming proof‑of‑concepts into production‑ready applications while enforcing data‑handling, safety, and observability controls.
Expactations:
• Architect secure, scalable AI services that meet business and regulatory requirements.
• Lead cross‑functional teams to prototype, validate, and roll out LLM‑based solutions.
• Maintain robust governance, CI/CD, and monitoring pipelines for continuous model improvement.
Key Responsibilities:
- Integrate and operationalize advanced AI models (LLMs, SLMs, LoRA fine‑tuning) into governed production stacks.
- Enforce data‑handling policies via code (prompt redaction, retrieval allow‑lists, per‑use‑case rules).
- Build Prompt/Agent CI/CD pipelines with evaluation gates, feature flags, canary deployments, and automated drift rollback.
- Manage model lifecycle: fine‑tuning, synthetic augmentation, registry lineage, and preservation of consented datasets.
- Implement observability: tracing (request → docs → output → tool calls), latency, cost SLOs, and alerting for hallucinations or safety incidents.
- Abstract cloud provider APIs (OpenAI, Gemini, Azure OpenAI, Vertex) behind a unified interface with routing, quota, and version control.
- Develop cloud‑native services (Lambda, CloudFront, S3), microservices, SOA, and EDA architectures using AWS (or Azure/Google/IBM Cloud).
- Create responsive web, web service, and batch applications using Java/J2EE, Maven, Ant, and related tools.
- Enforce audit‑ready lineage and retention for prompts, models, and data handling.
- Collaborate with stakeholders to document, explain, and approve AI concepts and changes.
Required Skills:
- Advanced proficiency in Python, PyTorch/TensorFlow, and Java.
- Experience with supervised, unsupervised, and reinforcement learning; NLP (tokenization, embeddings, semantic search).
- Hands‑on with LLM tools: LangChain, LangGraph, LlamaIndex, OpenAI APIs, MCP; Prompt engineering.
- Cloud architecture: AWS (Lambda, CloudFront, S3), Azure, Google Cloud, or IBM Cloud; Kubernetes fundamentals.
- CI/CD, MLOps, and monitoring experience (GitOps, TensorBoard, Prometheus).
- Data governance: PII/PCI/GDPR policies, redaction, masking, tenant isolation.
- API integration: OpenAI/ Gemini SDKs, fallback logic, request/response schema design.
- Strong communication for cross‑functional collaboration and stakeholder education.
Required Education & Certifications:
- Bachelor’s (or Master’s) degree in Computer Science, Data Science, Electrical Engineering, or related field.
- AWS Certified Solutions Architect – Associate/Professional or equivalent cloud certification.
- Additional certifications in Machine Learning, NLP, or MLOps are advantageous.