Senior Machine Learning Engineer: Search Quality

Constructor
UK
2 days ago

Role details

Contract type
Permanent contract
Employment type
Full-time (> 32 hours)
Experience level
Expert
Experience required
4 years minimum
Compensation
£80,000.0 - £120,000.0
Working hours
Regular working hours
Languages
English

Tech stack

A/B Testing Airflow Information Engineering Information Retrieval Python (Programming Language) Machine Learning Performance Tuning Recommender Systems SQL Databases Pytorch Large Language Models Apache Spark
+2 more
Low Latency Data Pipelines

Job description

As a Senior Machine Learning Engineer in the Search Quality team, you will improve the e-commerce experience for hundreds of millions of users across the world by building the systems that power relevance for global retailers - from fashion and grocery to electronics and hardware.

The mission is to measure search quality, push it higher, and catch degradations before the user does. You will achieve this through a blend of fine-tuned LLMs for relevance judgment, real-time models, and deep offline analysis of query logs.

Your primary focus will be relevance evaluation and quality improvements:

  • LLM-based evaluation. We fine-tune our own models to assess relevance. This involves teaching the model to understand query intent, represent items from messy catalog data, and align model judgments with real user behavior.
  • Real-time quality in production. Reranking, filtering, signal computation. Latency is a strict requirement, so quality vs speed tradeoff is constant.
  • Automated quality monitoring and agentic insights. Pipelines to detect degradations and find underperforming patterns. Agent-based systems that generate actionable recommendations for the product data and search configurations.

  • Multi-domain, multi-language, at scale - 40+ languages, 20+ domains. The models need to generalize across all of them - without per-customer rules or overrides.
  • No universal ground truth. A grocery retailer and a fashion retailer may have different perceptions on what “relevant” means.
  • Efficiency at scale. Optimizing and scaling LLM inference across our entire customer base., * Work with smart and empathetic people who will help you grow and make a meaningful impact.
  • Regular team offsite events to connect and collaborate.
  • Fully remote team - choose where you live.
  • ️ Unlimited vacation time - we strongly encourage all of our employees take at least 3 weeks per year.
  • ️ Work from home stipend! We want you to have the resources you need to set up your home office.
  • Apple laptops provided for new employees.
  • Training and development budget for every employee, refreshed each year.
  • Maternity & Paternity leave for qualified employees.
  • Base salary: $80k-$120K USD, depending on knowledge, skills, experience, and interview results
  • Stock options - offered in addition to the base salary

Requirements

  • Experience with search, information retrieval, or recommendation systems
  • Hands-on experience with fine-tuning, evaluation frameworks, and scaling LLM deployments
  • Strong Python and PyTorch. Fluency in SQL and data orchestration tools (Spark, Airflow)
  • Experience designing and running A/B tests to validate model impact
  • Excellent English communication skills
  • Experience collaborating in cross-functional teams (ranking, product, data engineering)

About the company

Launched in 2019, Constructor is an AI-first ecommerce search and discovery platform that helps shoppers find the right products at the right time and enables leading global e-commerce brands to drive meaningful revenue and conversion gains.

Apply for this position

This job is hosted externally. Click below to view the full posting and apply.

Apply on www.adzuna.co.uk

Good distractions

Talks and stories from around this role — technically off-topic, practically not.

54 sec

Generating multiple hook options for outreach A/B testing

Leandro Gomes da Silva Leandro Gomes da Silva · World Congress 2025

2:35 min

Preventing remote code execution in PyTorch models

Balázs Kiss · World Congress 2023

2:15 min

Empowering domain teams with an open data platform

Sandhya Menon Sandhya Menon · World Congress 2026 Europe

3:22 min

Creating dedicated AI assistants for advanced candidate sourcing

José Kadlec José Kadlec · World Congress 2025

2:24 min

Building scalable icon and button libraries internally

Nathalia Rus · World Congress 2022

1:06 min

Compiling PyTorch environments for advanced time forecasting

Christoph Lohrmann Christoph Lohrmann +1 · World Congress 2026 Europe

Videos

See all

Related articles

See all