AI Developer

ANRIC IT   |   Remuneration: $75,000 – $95,000 + Superannuation

Location: Melbourne, VIC (on-site)

Position: Full-time

Position Overview

ANRIC IT is seeking a skilled AI Developer to design, build, and deploy production-grade AI systems across the full engineering lifecycle. Reporting to the Development Lead, the successful candidate will take end-to-end ownership of AI-driven products — spanning LLM agent development, model fine-tuning, real-time pipelines, retrieval systems, and the cloud infrastructure that supports them.

This is a hands-on engineering role for a strong Python developer with demonstrated experience taking machine learning and large language model systems from concept through to production. The successful candidate will work on commercially significant projects within a collaborative, technically rigorous environment.

Key Highlights

•     End-to-End Ownership — Full responsibility for the product lifecycle, from architecture and prototyping through to deployment, evaluation, and ongoing optimisation of production AI systems.

•     Applied AI Engineering — Hands-on development of LLM agents, fine-tuned models, real-time speech and retrieval pipelines, and the supporting backend and infrastructure.

•     Production Impact — Building reliable, low-latency AI products that directly serve clients and deliver measurable commercial value.

Position Description

Role Purpose

The AI Developer is accountable for the design, development, evaluation, and deployment of production AI applications. Working across the entire stack — from telephony and speech pipelines through to LLM agent logic, retrieval, and cloud infrastructure — the role converts business requirements into reliable, scalable, production-ready systems.

Key Accountabilities

•     Design and implement LLM agent architectures, including tool-calling agents, retrieval-augmented generation (RAG), and structured output enforcement with validation and retry handling.

•     Fine-tune and optimise language models (e.g. QLoRA / PEFT) for structured extraction and domain-specific tasks, including quantisation, adapter management, and serving on constrained hardware.

•     Build and maintain evaluation, benchmarking, and A/B testing pipelines — including LLM-as-judge scoring, golden-dataset regression testing, and latency/throughput measurement under concurrency.

•     Develop real-time and streaming systems, including speech pipelines (VAD, STT, TTS) and low-latency response handling where applicable.

•     Implement retrieval systems using vector databases, embedding models, hybrid search, and cross-encoder reranking, with multi-format document ingestion.

•     Build robust, well-tested async Python backends (FastAPI) with clean, layered architecture, and supporting frontends (React) where required.

•     Provision, deploy, and manage cloud and GPU infrastructure, including reverse proxies, SSL, service management, and deployment automation.

•     Write clean, maintainable, well-documented code backed by comprehensive automated testing.

•     Maintain currency with emerging AI tools, models, and best practices, applying them where commercially appropriate.

Selection Criteria

Essential

•     Expert-level Python, including modern async/await, FastAPI, Pydantic, and pytest.

•     Demonstrated experience developing LLM-based applications: agent architectures, RAG, prompt engineering, and structured output enforcement.

•     Hands-on experience with the machine learning stack: PyTorch, Hugging Face (transformers / PEFT / TRL), and sentence-transformers.

•     Experience fine-tuning language models (e.g. QLoRA / LoRA), including quantisation and serving (Ollama, vLLM, or equivalent).

•     Experience building LLM evaluation and benchmarking systems (LLM-as-judge, golden datasets, regression and A/B testing).

•     Experience with retrieval systems and vector databases (e.g. Qdrant), embeddings, hybrid search, and cross-encoder reranking.

•     Sound understanding of end-to-end product development, from architecture and prototyping through to production deployment.

•     Experience deploying and managing cloud infrastructure (AWS, GPU instances), including nginx, systemd, Docker, Linux administration, and SSL/TLS.

•     Proven ability to deliver well-tested, production-grade systems and to work both autonomously and within a team.

•     Effective written and verbal communication skills.

Desirable

•     Experience with real-time speech and audio: Whisper STT, neural TTS, VAD, streaming audio, and voice cloning.

•     Telephony and protocol experience: SIP / RTP / SDP, G.711 codecs, or WebSocket-based media streaming.

•     Experience building Model Context Protocol (MCP) servers or custom AI agent / coding-agent harnesses and tooling.

•     Data engineering experience: ingestion pipelines, schema normalisation, and training-data preparation for embedding or model fine-tuning.

•     Front-end / full-stack experience with React and Vite for live, WebSocket-driven interfaces and dashboards.

•     Experience with NLP pipelines, sentiment analysis, or other applied ML domains.

•     A tertiary qualification in Computer Science, Software Engineering, Machine Learning, Data Science, or a related discipline, or equivalent practical experience.

Remuneration & Benefits

•     Competitive remuneration of $75,000 – $95,000 plus superannuation.

•     Hands-on involvement in commercially significant, cutting-edge AI projects.

•     Structured mentorship and clearly defined career progression pathways.

•     A professional, collaborative, and technically rigorous organisational culture.