Lead AI/ML Engineer, Algorithms & Research

Key Skills

Aartificial intelligence, Design, Large Language Models - LLMs, Leadership Skill, Machine learning techniques, NeurIPS

Job Description

The Algorithms and Research Team within the ML & AI organization is responsible for building the foundational models that support all vertical teams, including Search & Recommendations and Upwork’s latest initiative, developing Uma, our AI assistant.

As a Lead Machine Learning Engineer/Scientist, you will play a crucial role in developing and maintaining discriminative and generative AI models. You will design, train/finetune, and evaluate information retrieval systems capable of handling both structured and unstructured databases. Leveraging historical context and diverse data, you will create innovative models to drive significant business impact.

Responsibilities:

• Design and develop retrieval augmentation methods for structured, unstructured, and relational databases.

• Develop and utilize knowledge graphs for information retrieval. Orchestrate information retrieval using knowledge graphs in the loop.

• Investigate model explainability and work with trust and safety teams to enhance reliability and reduce hallucination in Gen-AI models.

• Collaborate cross-functionally with machine learning engineers to take models from research to production.

• Stay current with the latest advancements in Gen-AI and integrate relevant innovations into our systems.

• Mentor junior engineers, conduct code reviews, and enforce engineering best practices. Deliver high-quality solutions that improve overall efficiency.

What it takes to catch our eye:

• Proven experience in designing and developing information retrieval or RAG systems.

• Expertise in vector and graph databases.

• Expertise in Graph Neural Networks (GNNs) and related concepts.

• Familiarity with knowledge graphs and ontology engineering.

• Familiarity with search and recommendations in two-sided marketplaces.

• Passion for developing autonomous agentic systems using LLMs.

• Strong publication record in top-tier conferences like NeurIPS, CVPR, ICML, ICLR, etc.