AI Software Engineer

Numerator is looking for a passionate AI Software Engineer to join their growing Machine Learning team. In this role, you will design, build, deploy, and support high performance deep learning systems, making an immediate impact by expanding their technology platforms across various software products.

Responsibilities:

• Apply agentic LLM patterns (tool use, multi-step reasoning, orchestration) to automate high-complexity tasks that previously required human judgment
• Design and build GenAI-powered solutions for complex NLP tasks — NER, classification, information retrieval, summarization, and structured output generation
• Build and maintain production ML APIs and data pipelines that deliver model-driven automation and insight into Numerator's products at scale
• Work closely with MLOps engineers, product & engineering managers, and other teams, owning a large part of the process from problem understanding to shipping the solution
• Stay current with the fast-moving GenAI landscape and translate new capabilities into practical team impact

Requirements:

• 2+ years of experience building and deploying ML or GenAI systems in production
• Strong practical GenAI fundamentals: LLM APIs, context engineering, RAG, tool/function calling, skills, agents and evaluation methodology — you understand why these techniques work, not just how to call them
• Data acumen: you can critically assess a dataset, spot distribution problems, and reason about how data quality affects downstream model and business outcomes
• Product orientation: you engage with business context naturally, ask the right clarifying questions, and translate ambiguous requirements into well-scoped technical solutions
• Solid Python and software engineering fundamentals — clean, testable code, REST API design, debugging, and familiarity with CI/CD
• A genuine habit of self-improvement — you follow the field actively, experiment with new models and tools, and bring what's relevant back to the team
• Experience with agentic orchestration frameworks
• Fine-tuning experience with modern techniques — especially applied to domain adaptation for NLP tasks
• PyTorch or Hugging Face familiarity
• Familiarity with LLM evaluation frameworks and a structured approach to measuring model quality
• Inference optimization awareness — understanding latency/cost/accuracy tradeoffs for LLM solutions
• Experience building and deploying robust machine learning APIs in production environments (ideally cloud-based environments such as AWS or GCP)