[Hiring] Senior AI Solutions Architect @D2B

Role Description

This Senior AI Solutions Architect / CTO has strong relational database modelling, metadata-driven SaaS architecture, and LLM / agentic workflow experience. We are looking for a very senior, hands-on AI-native technical leader who can operate across architecture, solution design, code delivery, code review, and AI-assisted development. This person should be strong enough to contribute to the initial build directly, but also senior enough to help define the engineering standards, architecture patterns, and hiring bar for future developers. This is not a pure architect role and not a pure developer role. We need someone who can bridge both.

Key Responsibilities

• Understand and contribute to the Ellavait CRM and BI product architecture.
• Work with Micado leadership to shape the technical roadmap.
• Translate product requirements into secure, scalable, maintainable technical solutions.
• Use AI coding tools such as Claude Code, Cursor, GitHub Copilot, Codex, Replit, or similar to accelerate delivery.
• Personally contribute to development where required, especially in the early stages.
• Review AI-generated code critically before it reaches production.
• Protect the integrity of Ellavait’s multi-tenant architecture, including tenant isolation, role-based access control, data security, auditability, and maintainability.
• Help define coding standards, architecture decision records, delivery patterns, and AI development workflows.
• Support sprint planning by breaking larger initiatives into technically clear work packages.
• Help interview and assess future AI developers and QA automation resources.
• Act as a trusted technical counterpart to our internal leadership team.

Qualifications

• Senior full-stack software engineering background.
• Strong relational database design and data modelling experience, including complex entity relationships, lookup relationships, junction objects, field types, referential integrity, indexing, and query performance.
• Strong understanding of CRM-style data models, including accounts, contacts, leads, opportunities, activities, tasks, notes, emails, custom objects, and related records.
• Ability to design metadata-driven data structures where customers can configure objects, fields, layouts, relationships, and workflows without custom code.
• Strong understanding of tenant-scoped data access and the risks of data leakage in a multi-tenant SaaS platform.
• Strong understanding of AI and LLM application architecture, not only AI-assisted coding.
• Experience or strong working knowledge of agentic workflows, tool calling, retrieval augmented generation, vector databases, embeddings, structured outputs, and AI guardrails.
• Ability to design AI features that can safely reason over CRM records, relationships, documents, emails, tasks, and business workflows.
• Ability to design human-in-the-loop controls, audit logs, permission checks, and approval flows for AI-driven actions.

Preferred Technical Experience

• React and TypeScript.
• Python and FastAPI.
• PostgreSQL or other advanced relational databases.
• Advanced SQL, query optimisation, indexing, joins, and database performance tuning.
• Relational database design, including one-to-many, many-to-many, lookup relationships, junction objects, referential integrity, and schema migrations.
• Data modelling for CRM, ERP, workflow, BI, or metadata-driven applications.
• Metadata-driven SaaS architecture where objects, fields, layouts, relationships, permissions, and workflows can be configured dynamically.
• Multi-tenant SaaS platforms, including tenant isolation, tenant-scoped queries, shared versus dedicated data models, and data leakage prevention.
• Authentication and authorisation frameworks.
• Role-based access control and permission-aware application design.
• API architecture and integration design.
• Data pipelines and integration architecture.
• Event-driven architecture and background job processing.
• Automated testing, CI/CD, and release governance.
• Cloud deployment experience.
• Vector databases or vector search, such as pgvector, Pinecone, Weaviate, Qdrant, or similar.
• Embedding models and semantic search.
• LLM orchestration frameworks or patterns.
• AI agent frameworks, tool/function calling, and workflow automation.
• Retrieval-augmented generation architecture.
• Secure AI application design, including tenant-aware retrieval, permission checks, auditability, human-in-the-loop controls, and prevention of cross-tenant data exposure.