We are seeking a versatile, high-caliber AI/ML Engineer to drive the development of intelligent features across our product ecosystem. In this role, you won't just be training models in isolation—you will design and deploy end-to-end intelligent systems. Your work will span across building large-scale recommendation systems using graph databases, designing autonomous Reinforcement Learning (RL) agents capable of complex decision-making, and leveraging advanced NLP/LLM architectures.
The ideal candidate is an expert in the Python ecosystem who bridges the gap between sophisticated algorithmic design and practical software engineering—comfortable with everything from modern data scraping to fine-tuning open-source models.
Technical Stack & Ecosystem This is a Python-centric role. Experience or strong familiarity with the following libraries and frameworks is highly preferred:
Core Language: Python (Expert-level) and standard data stacks (NumPy, Pandas)
Graph Infrastructure: Neo4j or NebulaGraph
Machine & Deep Learning: PyTorch, TensorFlow, Scikit-Learn
Reinforcement Learning: Stable-Baselines3, Ray/RLLib, or custom RL environment design
Data Gathering: Playwright (preferred), Scrapy, BeautifulSoup LLM & Vector Infra: Hugging Face Transformers, vLLM, Llama/Qwen architectures, and Vector Databases (Chroma, Milvus, Pinecone, etc.)
Key Responsibilities
Graph-Based Recommendation Systems: Architect, optimize, and maintain scalable recommendation engines leveraging relational data and user behaviors within graph databases (Neo4j / NebulaGraph).
Autonomous Agent & RL Development: Design, train, and simulate Reinforcement Learning (RL) environments and multi-agent systems capable of autonomous decision-making, behavioral modeling, and strategic execution.
Modern Data Gathering: Own your data lifecycle. Build resilient scraping scripts using Playwright to harvest data from complex, dynamic web applications, and implement pipelines for data cleaning and tokenization.
Advanced LLM Implementations: Move beyond standard API integrations. Optimize vector spaces for retrieval-augmented generation (RAG), craft advanced prompt topologies, and execute custom fine-tuning workflows on open-source LLMs.
Production Deployment: Wrap, test, and run your models locally and prepare them for deployment in containerized production environments.
Required Skills & Qualifications
Core Machine Learning & Data Engineering
2+ years of professional experience delivering ML/DL solutions into production environments.
Graph Analytics: Hands-on experience structuring, querying, and optimizing complex relational data structures inside Neo4j or NebulaGraph.
Deep & Reinforcement Learning: Solid theoretical and practical grasp of DL architectures and RL paradigms (e.g., policy gradients, Q-learning, actor-critic frameworks). Modern Scraping Skills: Proven capability in building automated data-gathering pipelines, specifically handling dynamic, heavy-JavaScript web pages using automation frameworks like Playwright.
Avan Dade Pardaz is a fast-growing software company with a team of nearly 60 talented developers. We work closely together like a family while staying incredibly hardcore and passionate about what we do. We’re a team of curious nerds building cutting-edge solutions, constantly experimenting, learning, and pushing the limits of modern technology