Lead Software Solution Architect

Description:

We are hiring Lead Software Solution Architects whose mission is to architect and build enterprise software solutions, agentic applications, AI assistants, and intelligent platforms, while holding every solution to rigorous software architecture principles. This is a builder-architect role: you design the system and you ship it. AI acceleration multiplies your output; architectural discipline is what keeps that output enterprise-grade.

Key Responsibilities

Architect end-to-end software solutions: system decomposition, service boundaries, API contracts, data flows, security model, and deployment topology — then lead hands-on implementation to production.
Enforce architecture principles in real codebases: separation of concerns, dependency discipline, testability, and evolutionary design — through ADRs, design reviews, and code reviews.
Produce architecture deliverables clients act on: current/target state assessments, reference architectures, C4-style diagrams, and decision records.
Mentor engineers, raise the team's engineering bar, and contribute technical solutions and estimates to proposals.
Engage technical and enterprise stakeholders: design workshops, technical presentations, and demos.
Build with AI acceleration as the default workflow: spec-driven and agentic development with Claude/Claude Code, with full engineering ownership of everything shipped — reviewed, tested, verified.
Must-Have: Languages & Core Engineering

Python: Expert, production-grade: clean code, typing, packaging, async patterns, profiling, and test discipline (pytest, unit/integration/contract tests).
TypeScript/JavaScript: Working proficiency for full-stack AI products (Node.js, or similar) to architect and review across the stack.
Frontend frameworks: Angular (preferred) or React — component architecture, state management, and API integration for building product frontends; sufficient depth to architect and review full-stack solutions.
Development environments & AI-native tooling: Visual Studio and VS Code as base IDEs; Claude Code, Cursor, or GitHub Copilot (agent mode) as daily instruments for agentic, spec-driven development.
Data & persistence: Strong SQL and relational design (PostgreSQL/SQL Server); working knowledge of NoSQL and caching (MongoDB/Redis).
Platforms & DevOps: Git with disciplined branching and PR review workflows; CI/CD on Azure DevOps or GitHub Actions; Docker required, Kubernetes strongly preferred; Microsoft Azure; observability (logging, metrics, tracing — e.g., Prometheus/Grafana or Azure Monitor).
Enterprise & government integration: GSB, ESB/middleware (MuleSoft, Azure APIM, WSO2), SOAP/WSDL alongside REST; API gateways, OAuth2/mTLS, throttling, and integration with externally-owned services.
Must-Have: Software Architecture Principles — Applied, Not Recited

Design foundations: SOLID, separation of concerns, design patterns, clean/hexagonal architecture, domain-driven design fundamentals.
Distributed systems: microservices and modular monolith trade-offs, REST/gRPC API design, event-driven architecture and messaging (Kafka or equivalent), 12-factor applications.
Non-functionals by design: security (authN/authZ, OWASP, secrets management), scalability, resilience patterns (retries, circuit breakers, idempotency), performance, and cost.
Architecture communication: C4 diagrams, Architecture Decision Records, and trade-off analysis that stands up to challenge.
Must-Have: AI Engineering Stack

LLM APIs: Claude/Anthropic and OpenAI/Azure OpenAI APIs in production: tool use/function calling, structured outputs, streaming, context management, token/cost engineering.
Agentic frameworks: Multi-agent design and orchestration with LangGraph/LangChain or equivalent; MCP-based tool integration an advantage.
RAG: Retrieval architecture: chunking/embedding strategy, vector databases (pgvector, Qdrant, Pinecone, or Azure AI Search), hybrid retrieval, reranking.
AI quality & safety: Evaluation frameworks (golden sets, LLM-as-judge, regression suites), guardrails, hallucination controls, and production monitoring of AI behavior.
AI-accelerated development: Claude Code (or equivalent) as a daily engineering instrument: prompt/spec decomposition, agentic coding workflows, and rigorous review of generated output — “the AI wrote it” is never an explanation for a defect.
The Skills That Make an Architect — What We Assess For

Trade-off articulation: you defend why, not just describe what — and can argue the option you rejected.
Judgment under acceleration: knowing when AI-generated design or code is good enough, and when it violates the architecture — catching it before production does.
Technical leadership: standards-setting, constructive review of others' work, and mentoring.
Client-grade communication: explaining architecture decisions to engineers and to ministerial-level stakeholders with equal credibility.
Experience & Qualifications

6+ years in software engineering, including 1+ years building LLM-based software in production; proven enterprise-scale, end-to-end delivery.
Government or large-enterprise client experience; consulting environment strongly preferred; GCC an advantage.
3+ years managing engineering teams with accountability for team output — including performance management, delivery cadences, and a track record of developing engineers toward promotion.
Bachelor's in Computer Science/Engineering or related; relevant certifications desirable; Arabic speakers preferable.