Software Architect (Python, AWS, AI)

Software Architect (Python, AWS, AI)

You'll architect a sophisticated, multi-tenant AI security platform —a 7-stage LLM-powered control-generation pipeline that generates right-sized cybersecurity controls with full bidirectional traceability. This is not a proof-of-concept; it's a production system where reproducibility, security, and audit-ability are foundational.

You own the high-level design decisions: where the deterministic/AI boundary lives, how to enforce multi-tenant isolation at the database layer, how to defend against prompt injection, how to validate LLM outputs at scale. You'll guide a small team of AI engineers, baseline architectural decisions early, and remain the escalation point for ambiguous calls.

Key Responsibilities

• Act as the architecture authority – own the deterministic/AI boundary design, multi-tenant data isolation model, and LLM security posture; baseline architectural decisions in week 1 that lock the team's constraints
• Technical owner from design to delivery – tailor solutions to Alpine's stated requirements (FastAPI + PostgreSQL + AWS Bedrock + ECS Fargate) while defending against scope creep and technical debt
• Partner with stakeholders to capture requirements – work with client leadership to translate business and compliance needs into architectural constraints (e.g., multi-tenant isolation, run-to-run consistency guarantees, audit trails)
• Assess and plan for technology insertion – evaluate architectural trade-offs (e.g., why a single bounded AI step beats two, why RLS + FORCE ROW LEVEL SECURITY beats application-layer filtering, why Pydantic schema validation is the output gate)
• Manage risk identification and mitigation – surface and mitigate delivery risks (e.g., CISO golden-set availability, prompt-injection surface area, cross-tenant isolation defects) with concrete mitigation strategies
• Influence and communicate vision – define the long-term product roadmap (Tools 1–5), technical vision (the 5-section Source of Truth structure, the 7-stage pipeline design), and development strategy (agentic-orchestration approach, vertical slices, always-functional milestones)
• Contribute to code reviews and architectural artifacts – review critical paths (Bedrock adapter, RLS policies, Stage 4 calibration logic); document architectural decision records (ADRs) that capture the why, not just the what
• Lead and mentor the team – guide the AI/LLM Lead and Backend engineers on AI reliability, security patterns, and multi-tenant design; escalate ambiguity upward rather than letting it flow down to individual engineers

Minimum Qualifications

• Bachelor's degree in Computer Science or related field (or equivalent professional experience)
• 5+ years as a Software Architect with hands-on expertise in: Python and async patterns (FastAPI, async/await); PostgreSQL and multi-tenant database design (RLS, row-level security, schema isolation); AWS cloud architecture (Fargate, RDS, IAM, Secrets Manager, Bedrock); LLM system design (prompt engineering, consistency guarantees, output validation, prompt injection defense); Security architecture (least-privilege access, encryption in transit, audit logging, cross-tenant isolation); Operational excellence in design methodologies – fluency in architectural patterns (layered,