Pantheion AI

AI Engineer — The Forge (Open-Weight Platform)

Abu Dhabi, UAEFull-TimeOn-sitePosted Apr 6, 2026

LLM Fine-tuningLoRAPythonDockerKubernetesDeveloper Tooling

Role overview

The Forge's success depends on making Pantheion Sovereign Weights genuinely easy for GCC startups to pull, deploy, and fine-tune in their own private cloud environments. As AI Engineer for The Forge, you will build the model tooling, fine-tuning infrastructure, and deployment systems that sit between Pantheion AI's trained model weights and the Arabic AI startup ecosystem. This is a product-facing engineering role — your output is what startups touch first, and their experience will determine how rapidly the Pantheion developer ecosystem grows.

What you will do

Build and maintain the Sovereign Weights distribution infrastructure — model packaging, secure download pipelines, weight integrity verification, and version management for open-weight releases
Develop the fine-tuning tooling suite for The Forge: one-click fine-tuning scripts, LoRA/QLoRA adapters, dataset formatting tools, and evaluation pipelines optimised for Arabic language tasks and GCC enterprise use cases
Build private cloud deployment packages for Pantheion model variants: Docker containers, Kubernetes Helm charts, inference server configurations (vLLM, Ollama, TGI), and cloud-agnostic deployment guides
Develop and maintain the Arabic Developer Platform SDK: Python client library, REST API wrappers, streaming inference utilities, and code examples covering common GCC startup use cases
Design quantization and model compression pipelines that produce smaller, faster Forge variants suitable for startup infrastructure budgets without unacceptable Arabic language performance degradation
Build evaluation tooling that allows startups to benchmark their fine-tuned Pantheion variants against standard Arabic NLP benchmarks before production deployment
Engage directly with developer community feedback — triaging issues, writing guides, running developer workshops, and ensuring the technical experience of The Forge is world-class
Monitor and analyse model usage patterns across Forge deployments to inform future model release priorities and tooling investment

Skills profile

Required skills

LLM Fine-tuningLoRAPythonDockerKubernetesDeveloper Tooling

Required qualifications

Domain knowledge

4+ years of AI/ML engineering experience with strong hands-on background in LLM fine-tuning, deployment, and inference optimization
Deep expertise in parameter-efficient fine-tuning methods: LoRA, QLoRA, adapter tuning, and prompt tuning
Experience building and maintaining developer-facing AI tooling — SDKs, CLIs, notebooks, or open-source ML libraries
Strong proficiency in Python and familiarity with inference serving frameworks (vLLM, TGI, Ollama, or equivalent)
Experience with containerization and cloud-agnostic deployment (Docker, Kubernetes, major cloud providers)
Excellent technical communication: able to write clear documentation, tutorials, and code examples for an external developer audience

Preferred qualifications

Bonus domain experience

Prior contributions to open-source LLM tooling ecosystems (HuggingFace, LangChain, LlamaIndex, Axolotl, or equivalent)
Experience with Arabic NLP fine-tuning — data preparation, tokenizer adaptation, and evaluation for Arabic language tasks
Familiarity with model quantization frameworks (GPTQ, AWQ, GGUF) and their tradeoffs for Arabic language models
Prior experience in developer relations, developer advocacy, or community engineering roles
Understanding of GCC startup technical infrastructure patterns and cloud provider preferences in the region