Chief Innovation Architect

Job Title: Lead Technical Architect

We are seeking an experienced and skilled technical leader to oversee the end-to-end design and delivery of AI-powered solutions.

Main Responsibilities:

• Define Reference Architecture: Develop a comprehensive architecture for ingestion, lakehouse, feature/vectors, models, APIs/UX; make informed decisions on build/buy strategies.
• NLP and GenAI/RAG Implementation: Select, train, and operationalize NLP, anomaly/fraud models, and GenAI/RAG components; establish human-in-the-loop feedback mechanisms.
• Experiment Tracking and Model Registry: Implement experiment tracking, model registry, CI/CD for models, automated evaluation, drift monitoring, and rollback procedures.
• Pipeline Design and Guardrails: Design retrieval pipelines (chunking, embeddings, vector namespaces), guardrails (prompt policies, allow-lists, PII redaction), and citation-based responses.
• Feature Store and Data Quality: Oversee feature store, labelling strategy, high-quality gold datasets, enforce DQ rules, and lineage management.
• Performance Optimization: Right-size SKUs, caching/batching, cost/per-token dashboards, SLOs for latency/throughput.
• Technical Leadership: Break down epics/stories, estimate and sequence work, unblock the team, run technical design reviews.
• Business Policy Translation: Translate business policy into rule + ML ensembles, present options, risks, and trade-offs.
• Testing and Observability: Establish testing pyramid (unit, data, model evals, e2e), performance targets, observability dashboards.
• Mentorship and Standards: Produce design docs, runbooks, SOPs; mentor engineers; uplift coding and review standards.

Requirements:

• 10+ years of software development experience with at least 5 years leading AI/ML projects.
• Supervised/unsupervised modelling, anomaly detection, NLP (extraction, classification, NER), OCR pipelines; evaluation design (precision/recall, ROC/PR).
• LangChain/LangGraph; embeddings; vector databases / Azure Cognitive Search (vector); prompt engineering & safety patterns.
• MLflow, model registry, online/offline evals, data/version management; CI/CD for models.
• Delta Lake/Lakehouse (bronze/silver/gold), Azure Data Lake Gen2, Databricks/ADF, schema/versioning, Great Expectations/Deequ.
• Python (FastAPI), eventing (Event Grid/Service Bus), containerization (AKS/Azure Container Apps), APIM.
• Cognitive Search, Azure OpenAI, Key Vault, Monitor/App Insights, Purview, Cosmos DB/Azure SQL.
• Technical decision-making, cross-team coordination, stakeholder communication, and mentoring.

Preferred Skills:

• Certifications in AI/ML disciplines.
• Hands-on experience with explainable AI and AI governance.
• Familiarity with regulatory compliance standards for financial data (e.g., SOX, GDPR).

Qualifications:

• Bachelor's Degree in Computer Science or related science field or equivalent.

About This Role:

This is an exciting opportunity for a technical leader to join our team. We are looking for someone who has experience in leading AI/ML projects and can help us deliver innovative solutions.

What You Will Do:

You will be responsible for overseeing the end-to-end design and delivery of AI-powered solutions. This includes defining reference architecture, implementing NLP and GenAI/RAG components, and establishing experiment tracking and model registry.

Key Responsibilities:

1. Define Reference Architecture: Develop a comprehensive architecture for ingestion, lakehouse, feature/vectors, models, APIs/UX; make informed decisions on build/buy strategies.
2. NLP and GenAI/RAG Implementation: Select, train, and operationalize NLP, anomaly/fraud models, and GenAI/RAG components; establish human-in-the-loop feedback mechanisms.
3. Experiment Tracking and Model Registry: Implement experiment tracking, model registry, CI/CD for models, automated evaluation, drift monitoring, and rollback procedures.
4. Pipeline Design and Guardrails: Design retrieval pipelines (chunking, embeddings, vector namespaces), guardrails (prompt policies, allow-lists, PII redaction), and citation-based responses.
5. Feature Store and Data Quality: Oversee feature store, labelling strategy, high-quality gold datasets, enforce DQ rules, and lineage management.
6. Performance Optimization: Right-size SKUs, caching/batching, cost/per-token dashboards, SLOs for latency/throughput.
7. Technical Leadership: Break down epics/stories, estimate and sequence work, unblock the team, run technical design reviews.
8. Business Policy Translation: Translate business policy into rule + ML ensembles, present options, risks, and trade-offs.
9. Testing and Observability: Establish testing pyramid (unit, data, model evals, e2e), performance targets, observability dashboards.
10. Mentorship and Standards: Produce design docs, runbooks, SOPs; mentor engineers; uplift coding and review standards.