Senior AI Engineer: Lead LLM Apps & Production Systems

E-Merge; South Africa; On-site; ZAR 900,000 - 1,100,000; Full time

Job summary

e-Merge IT Recruitment is seeking a Senior AI Engineer based in Pretoria to lead the development of production-grade AI solutions using LangChain, LangGraph, and state-of-the-art LLM tooling. You will influence architectural decisions, optimize inference costs, and build scalable cloud-native applications.

The role requires expert Python, strong API design with FastAPI/Flask, and experience with vector databases and PostgreSQL.

Qualifications

• Expert proficiency in Python is mandatory.
• Deep architectural understanding and production experience with LangChain; LangGraph for multi-agent orchestration is desirable.
• Advanced knowledge of vector search optimization and indexing with Pinecone/Weaviate/Qdrant/pgvector.
• Understanding of fine-tuning principles and trade-offs vs RAG; managing context windows and optimizing inference costs/latency for GPTs/Claude/local LLMs.
• 6+ years of general Software Engineering experience.
• Demonstrated technical leadership in delivering complex software projects.
• Proven portfolio of building LLM-based applications or agents.
• Expert skills in FastAPI or Flask including asynchronous design patterns.
• PostgreSQL expertise; NoSQL and data consistency concepts.
• Extensive cloud experience (AWS Bedrock/SageMaker/Lambda/ECS) or Azure.
• Docker and orchestration (Kubernetes or ECS).
• IaC proficiency with Terraform, AWS CDK, or CloudFormation.

Skills

Python

LangChain

LangGraph

LLM applications

Fine-tuning concepts

RAG

API design

Education

IT Degree (Computer Science/Engineering)

Tools

Pinecone

Weaviate

Qdrant

pgvector

FastAPI

Flask

PostgreSQL

NoSQL

AWS Bedrock

SageMaker

Lambda

ECS

Docker

Kubernetes

Terraform

AWS CDK

CloudFormation

Job description

e-Merge IT Recruitment is seeking a Senior AI Engineer based in Pretoria to lead the development of production-grade AI solutions using LangChain, LangGraph, and state-of-the-art LLM tooling. You will influence architectural decisions, optimize inference costs, and build scalable cloud-native applications.

The role requires expert Python, strong API design with FastAPI/Flask, and experience with vector databases and PostgreSQL.