Senior AI Engineer & Founding Architect Multi-Agent

Avenga; Turkey; On-site; TRY 600,000 - 900,000; Full time

Job summary

Avenga is seeking a Senior AI Engineer / Founding Engineer to build end-to-end AI systems in a fast-paced R&D environment. The role covers data scraping, multi-agent backends, and a responsive 3D UI, with Python-based APIs and async workflows.

You will shape AI/ML designs, agent orchestration, vector databases, and RAG architectures, collaborating across geospatial datasets and external tools, while delivering scalable, robust systems in a distributed team.

Qualifications

• Experience leading ML design decisions and technical direction.
• Hands-on with frameworks like LangGraph/LangChain or similar.
• Proven ability to design multi-agent orchestration and state management.
• Strong Python and API development using FastAPI.
• Experience with vector databases and retrieval augmented generation.

Responsibilities

• Build end-to-end AI systems spanning data scraping/ingestion, multi-agent backends, and 3D UI rendering.
• Develop AI systems using LLMs, RAG architectures, vector databases, embeddings, semantic search, function calling, and external tools.
• Design scalable Python/FastAPI backends with asynchronous programming for long-running processes.
• Create structured and validated inputs and outputs for AI systems and APIs.
• Collaborate in an R&D/POC environment and adapt to changing priorities.
• Apply problem-solving and ownership to evolving AI/ML challenges.

Skills

ML Leadership

Agentic Orchestration

Multi-Agent Systems

GenAI & LLMs

RAG

Function Calling

Vector Databases

Python API

Async Programming

Data Modeling

Job description

Avenga is seeking a Senior AI Engineer / Founding Engineer to build end-to-end AI systems in a fast-paced R&D environment. The role covers data scraping, multi-agent backends, and a responsive 3D UI, with Python-based APIs and async workflows.

You will shape AI/ML designs, agent orchestration, vector databases, and RAG architectures, collaborating across geospatial datasets and external tools, while delivering scalable, robust systems in a distributed team.