AI Systems Architect - Agentic Solutions (Remote, Anywhere in Pakistan, USD Salary)

Requirements:

• 5 - 8 years of experience in software engineering, AI systems, or related fields
• Hands-on experience with open-source LLMs such as Llama and Mistral
• Proficiency with orchestration frameworks, including LangChain, AutoGen, and CrewAI
• Strong knowledge of on-premise model deployment and optimization using vLLM, TGI, and Ollama
• Deep understanding of agent architectures, including multi-agent systems, memory, and tool integration
• Proven experience with RAG pipelines, vector databases, and prompt engineering at scale
• Ability to navigate enterprise processes while maintaining innovation velocity
• Experience with Confluence, Jira, and Bitbucket or similar platforms
• Strong documentation and formal specification skills
• Ability to translate vague client needs into concrete technical solutions
• Skilled at communicating complex systems in simple terms
• Comfortable engaging stakeholders across both technical and business domains
• Collaborative mindset with the ability to thrive in cross-functional environments

Responsibilities:

• Design and implement autonomous agentic systems using open-source LLMs (Llama, Mistral) and orchestration frameworks (LangChain, AutoGen, CrewAI)
• Deploy, fine-tune, and optimize models on-premise using vLLM, TGI, and Ollama
• Build and integrate advanced agent features, including tool use, memory systems, and multi-agent coordination
• Develop and scale RAG pipelines, vector database solutions, and prompt engineering techniques
• Present architectural approaches and AI solutions to non-technical stakeholders in clear, understandable terms
• Document architectural decisions in Confluence, manage development sprints in Jira, and maintain clean, reviewable code in Bitbucket
• Create and defend technical specifications in formal review processes
• Balance rapid prototyping and innovation with governance and compliance requirements