Software Architect

AI Software Architect (On-Premise AI & High-Performance Computing)

Company Overview

AGENTIC AI ARTIFICIAL INTELLIGENCE DEVELOPING SERVICES is an AI engineering company based in Abu Dhabi, dedicated to building autonomous agent systems, enterprise workflow orchestrations, and high-performance, private on-premise AI platforms.

Role Overview We are seeking an experienced, hands-on Software Architect to lead the technical architecture, design, and deployment of enterprise backend systems and high-throughput AI workloads running on on-premise server infrastructure and dedicated GPU clusters. You will bridge production backend engineering with local AI models, fine-tuning pipelines, model serving, and output accuracy management.

Key Responsibilities

* Architect, develop, and maintain robust, production-grade backend services, APIs, and microservices using Python (FastAPI/Flask) and Node.js.

* Lead the deployment, orchestration, and lifecycle management of AI workloads on on-premise Linux servers and GPU infrastructure (e.g., bare-metal setups, NVIDIA DGX clusters).

* Host, fine-tune, and serve open-source foundation models utilizing the Hugging Face ecosystem, PyTorch, and optimized inference runtimes (e.g., vLLM, TensorRT, Triton Inference Server).

* Build end-to-end pipelines for model training, fine-tuning, evaluation, and accuracy management, implementing verification controls to reduce hallucinations.

* Design data architectures, caching layers, and vector search systems using PostgreSQL, Supabase, Redis, and Vector DBs (e.g., pgvector, Qdrant, Milvus).

* Design autonomous multi-agent systems, reasoning loops, and workflow automations using frameworks such as LangChain, LangGraph, LlamaIndex, or n8n.

* Implement containerization and cluster orchestration using Docker and Kubernetes optimized for local hardware.

* Establish system security, hardware resource allocation (GPU/CPU optimization), and observability (Prometheus, Grafana, OpenTelemetry).

Required Qualifications

* Minimum 5 years of professional experience in software engineering, backend engineering, AI applications, or related technical roles with hands-on production delivery.

* Deep programming expertise in Python and Node.js.

* Strong hands-on experience deploying, managing, and fine-tuning open-source models using Hugging Face in GPU-accelerated environments.

* Solid background in on-premise server administration, Linux environments, and local GPU cluster optimization.

* Practical knowledge of model accuracy measurement, output verification, latency optimization, and quantization.

* Strong experience with containerization (Docker/Kubernetes), microservices design, and database architectures (PostgreSQL, Supabase, Vector DBs).

* Practical familiarity with agentic frameworks (LangChain, LangGraph, LlamaIndex) and automation platforms (n8n).

Preferred Qualifications

* Direct experience with NVIDIA DGX systems, MLOps, model serving, and high-performance AI infrastructure.

* Experience building automations for enterprise systems, ERPs, or executive workflows.

* Ability to work effectively in an on-site business environment in Abu Dhabi

* Company: AGENTIC AI ARTIFICIAL INTELLIGENCE DEVELOPING SERVICES

* Workplace Type: On-site

* Location: Abu Dhabi, Abu Dhabi Emirate, United Arab Emirates

* Job Type: Full-time