MLOps & AI Architecture Engineer

About the role

Neural Surge AI designs and ships intelligent systems covering LLMs, retrieval (RAG), AI agents, machine learning models and data pipelines. It is looking for an MLOps and AI Architecture Engineer to shape how its proprietary AI pipeline is built, deployed, monitored and scaled. The role suits someone who has already put ML or LLM systems into production and wants more ownership; real responsibility starts from the first month.

Key responsibilities

• Design the architecture of the proprietary AI pipeline, connecting data, models, prompts, retrieval and agents
• Build and run deployment pipelines for LLM, RAG and ML systems, with CI/CD, versioning and safe rollbacks
• Set up monitoring for quality, latency, cost and failures
• Build evaluation frameworks and test sets so every model or prompt change can be proven to help
• Choose the right model, caching and scaling approach for each task
• Manage data and feature pipelines
• Manage cloud infrastructure, containers, secrets and environments (AWS preferred)
• Add guardrails, access controls and audit trails so AI outputs are safe, explainable and fair
• Review new models, vendors and tools and recommend what to adopt or replace
• Write architecture notes and runbooks for other engineers

Required profile

• 3 to 5 years in MLOps, ML engineering, DevOps or backend engineering, with at least 2 years deploying ML or LLM systems to production
• Ability to think in systems and explain why a design is reliable, scalable and cost-effective
• Ownership mindset: staying with a problem until it works in production

Required skills

• Docker and a major cloud platform (AWS preferred)
• CI/CD, monitoring, logging, APIs and databases
• Practical experience with LLM APIs (OpenAI, Anthropic or similar) and RAG
• Nice to have: Kubernetes, Terraform, MLflow, Airflow, vector databases, ONNX/TensorRT/vLLM/Triton, LangChain/LlamaIndex, Prometheus/Grafana/OpenTelemetry, Spark/Kafka, WebRTC, Urdu NLP, computer vision

The position is on-site in Lahore. Applications are made through LinkedIn or by sending a CV to the company's HR team.