company-logo
Snr Applied AI Architect

Description

A client of byteSpark.ai is seeking a Senior Applied AI Architect to lead the architecture and delivery of advanced AI solutions for high-volume telecommunications and government use cases. The successful candidate will define end-to-end architectures spanning large language models, natural language processing, speech recognition, computer vision, diffusion models, and multimodal analytics. This role will translate complex business, security, data, and performance requirements into documented system components, interfaces, deployment patterns, and implementation roadmaps. The architect will guide engineering teams through model selection, fine-tuning, evaluation, integration, optimization, and reliable production deployment. They will design scalable machine learning pipelines capable of processing real-time and batch voice, text, image, and data streams containing billions of data points. The position requires ownership of key technical decisions involving distributed training, inference infrastructure, MLOps, privacy, observability, resiliency, and cost-performance tradeoffs. The architect will collaborate with product, data, platform, security, and client stakeholders to integrate AI capabilities with existing applications, data sources, cloud platforms, and enterprise systems. They will establish standards for responsible AI, reproducibility, model monitoring, technical documentation, and measurable solution quality. The role also includes mentoring engineers, communicating architecture to technical and nontechnical audiences, and evaluating emerging AI methods that can deliver practical business and operational value.

Requirements

  • At least 7 years of relevant AI or machine learning experience, including senior ownership of production AI architecture and delivery across complex initiatives.
  • Expert-level Python skills and substantial hands-on experience with PyTorch, TensorFlow, JAX, Hugging Face Transformers, or comparable deep learning frameworks.
  • Demonstrated expertise architecting large language model and NLP solutions using transformer architectures such as GPT, BERT, T5, LLaMA, or comparable models.
  • Hands-on experience with speech and audio processing architectures, including ASR technologies such as Whisper or Wav2Vec2 and real-time or batch voice pipelines.
  • Proven ability to design and deploy end-to-end multimodal AI pipelines encompassing data ingestion, model training, evaluation, serving, integration, monitoring, and continuous improvement.
  • Strong knowledge of model adaptation and optimization techniques, including LoRA, QLoRA, instruction tuning, quantization, pruning, distillation, ONNX, or TensorRT.
  • Production experience with distributed computing and MLOps technologies such as Docker, Kubernetes, MLflow, Ray, Apache Spark, DVC, and cloud-based machine learning platforms.
  • Ability to translate business, security, scalability, and performance requirements into documented AI architectures, interfaces, implementation plans, and defensible technical decisions.

Desirable

  • Advanced degree in computer science, machine learning, data science, mathematics, or a closely related discipline.
  • Experience with diffusion models, synthetic data generation, multimodal content generation, VAEs, GANs, or related generative modeling methods.
  • Knowledge of responsible and privacy-preserving AI, including explainability, bias detection, federated learning, adversarial robustness, or secure government deployments.
  • Experience with real-time streaming inference, edge deployment, hardware-aware optimization, or high-volume telecommunications data processing.
  • Published AI research, open-source contributions, patents, conference presentations, or demonstrated leadership in advancing engineering practices.
Role Highlights
💰 Compensation
USD 80K-90K/year + benefits structure
📍 Location
Dubai
💼 Work Location Type
Hybrid
📄 Employment Type
Full-time
📈 Job Level
Senior
⌛ Experience
8+ years
🏢 Department
Information Technology
🏭 Industry
Technology, Information & Media
🔹 Sub-Industry
AI Development

⚠️

Applications for this role are closed!
Thank you for visiting!
Frequently Asked Questions (FAQs)Have a question? Check here for quick answers

This confidential recruitment process is being managed by byteSpark.ai on behalf of an established international technology organisation operating in highly specialised, mission-critical environments. byteSpark.ai is not the employer for this position. Further information about the company, engineering leadership, products, and AI roadmap will be shared with shortlisted candidates at the appropriate stage.

This is an opportunity to work on applied AI problems where reliability, performance, security, and real-world usability genuinely matter. The successful candidate will help shape AI capabilities involving Large Language Models, Agentic AI, NLP, speech, and multimodal technologies within a complex enterprise product environment. The role offers the chance to influence how AI is designed, integrated, and operationalised rather than simply consuming third-party models.

The successful candidate will design, build, fine-tune, evaluate, and deploy AI capabilities that can be integrated into real products and operational workflows. This includes developing LLM and machine-learning services, connecting models to APIs and enterprise data sources, improving inference performance, evaluating reliability, and working closely with engineering and product teams to move AI initiatives from concept into production.

The client is not building generic AI demonstrations. Its technology operates in environments where systems must be dependable, secure, explainable, and capable of handling complex data and workflows. Engineers will therefore be exposed to meaningful technical challenges involving model quality, latency, integration, deployment constraints, and production reliability.

Candidates will stand out if they can show that they have personally taken AI systems beyond experimentation. Strong evidence of production LLM applications, model fine-tuning, Agentic AI workflows, enterprise integration, PyTorch, Hugging Face, MLOps, optimisation, and measurable technical or commercial outcomes will significantly strengthen an application.

The client is looking for a hands-on engineer who can move confidently between model development, software engineering, experimentation, and production delivery. The ideal candidate is intellectually curious, technically rigorous, comfortable taking ownership, and motivated by solving difficult applied AI problems rather than working only on research or isolated prototypes.

The role is expected to offer meaningful technical ownership. The successful candidate will contribute to architecture decisions, model selection, experimentation, deployment approaches, and the development of reusable AI capabilities. This is not a narrow implementation role; the engineer will help shape how the organisation applies AI across its product environment.

The position is based in Portugal. Compensation will be aligned with the candidate's experience, technical depth, and overall fit for the role. Further details regarding salary, benefits, working arrangements, and any applicable relocation support will be discussed with shortlisted candidates during the recruitment process.