Teams at Newcode.ai

Own the systems and frameworks used to evaluate Newcode’s AI products. Build automated and human-in-the-loop evaluations, datasets, benchmarks, metrics, regression suites, monitoring, and quality tooling. Analyze model failures, investigate root causes, and evaluate changes to models, prompts, retrieval systems, agents, and workflows. Collaborate with Software Engineering, Product, and Legal Engineering to translate legal workflows into measurable quality standards and improvements.

Own and operate the end-to-end retrieval pipeline: parsing/chunking, embeddings, vector indexing, semantic+keyword search, reranking, and agentic retrieval loops. Build evaluation (golden sets, metrics, regression tests), ship robust Python/FastAPI services, ensure tenant isolation and data security, and make multilingual retrieval reliable under production load.

Lead testing strategy for AI-powered legal products: build evaluation frameworks, design test plans, measure AI outputs, run regression and UAT, identify risks, and partner with engineering to improve quality and scale the testing function.