Senior Machine Learning Engineer
BloomreachIndiaOn-Site
Needs a free Superjobs account. You'll finish on Bloomreach's Greenhouse page.
Skill score
Summary
A Senior Machine Learning Engineer on the Search Scaling team takes ML models behind Bloomreach search into production, building low-latency serving, query-time ML and large catalog data pipelines. It suits engineers with solid ML depth and strong backend skills who can own problems end to end and work with the Search Quality team.
- Function
- Data Science
- Estimated salary
- Est. ₹14–21 LPA
- Estimated for a senior role, ~5 yrs experience, India. Not published by the employer.
- Work mode and location
- On-Site, India
- Experience
- 5+ yrs
Job description
Overview
Bloomreach is seeking a Senior Machine Learning Engineer to join our Search Scaling team. The team takes the ML and AI behind Bloomreach search and makes it run in production: in the live search path that answers every shopper query, and in the data pipelines that prepare product catalogs for search.
Our search powers online stores for many merchants, each with its own catalog, vertical, language and region. The Scaling team's job is to make our search and discovery features work well for all of them, and to turn models into reliable, low-latency, cost-efficient systems at that scale.
We are looking for someone who combines solid ML expertise with strong software engineering skills and takes end-to-end ownership of technical problems. You will partner closely with our Search Quality team, which develops the relevance and ranking models behind our search.
Roles and responsibilities
- Design, build and operate ML components in the search serving path, within tight latency budgets.
- Build query-time ML that helps search understand and handle each query.
- Build and scale ML and LLM data pipelines across large product catalogs, with attention to cost and freshness.
- Tune models across merchants, verticals and languages.
- Build and scale online features and model serving infrastructure.
- Set up offline evaluation and safe rollout for production models: baselines, A/B tests, fallbacks, and latency and cost tracking.
- Improve observability and diagnosis of ML systems in production.
- Work with Search Quality engineers to bring their models into production within latency and cost targets.
- Contribute to technical design and influence decisions across teams.
Skills required
Top skills
- Python
- Go or Java
- PyTorch or TensorFlow
- ML model serving
All skills (7)
- LLM serving
- AWS or GCP
- Large-scale data pipelines
- Search ranking and query understanding
- Solr or Elasticsearch
- Vector search (Milvus, Qdrant)
- ML evaluation and observability tooling
Qualifications
- BE/B.Tech/ME/M.Tech in Computer Science or a related discipline, with a strong mathematical foundation and excellent programming skills in Python. Experience with a backend language such as Go or Java.
- 5+ years of experience building ML-driven systems in production.
- Strong software engineering skills, with hands-on experience serving ML models and LLMs in scalable cloud systems (AWS/GCP) under tight latency and cost budgets.
- Experience building large-scale data and ML pipelines.
- Strong problem-solving and analytical skills: able to break a problem down into its model-training and backend-engineering parts, and, during development, analyze results to pinpoint the cause of an issue and choose the right next step.
- Experience working across multiple customers, domains or languages.
- Working knowledge of search, such as query understanding, ranking, autosuggest or faceting.
- Experience with deep learning frameworks such as PyTorch or TensorFlow.
Nice to have
- Experience with Solr, Elasticsearch or a similar search engine.
- Experience with vector search in production, such as Milvus or Qdrant.
- Experience building evaluation, observability or explainability tooling for ML models, LLMs and serving systems.
Other details
• We have a People Development Program - participating in personal development workshops on various topics run by experts from inside the company. We are continuously developing & updating competency maps for select functions.
- Our resident communication coach Ivo Večeřa is available to help navigate work-related communications & decision-making challenges.*
- Our managers are strongly encouraged to participate in the Leader Development Program to develop in the areas we consider essential for any leader. The program includes regular comprehensive feedback, consultations with a coach and follow-up check-ins.
- Bloomreachers utilize the $1,500 professional education budget on an annual basis to purchase education products (books, courses, certifications, etc.)*
Well-being
• The Employee Assistance Program -- with counselors -- is available for non-work-related challenges.*
- Subscription to Calm - sleep and meditation app.*
- We organize ‘DisConnect’ days where Bloomreachers globally enjoy one additional day off each quarter, allowing us to unwind together and focus on activities away from the screen with our loved ones.
- We facilitate sports, yoga, and meditation opportunities for each other.
- Extended parental leave up to 26 calendar weeks for Primary Caregivers.*
Compensation
- Restricted Stock Units or Stock Options are granted depending on a team member’s role, seniority, and location.*
- Everyone gets to participate in the company's success through the company performance bonus.*
- We offer an employee referral bonus of up to $3,000!
- We reward & celebrate work anniversaries -- Bloomversaries!*
Excited? Join us and transform the future of commerce experiences!
If this position doesn't suit you, but you know someone who might be a great fit, share it - we will be very grateful!
Company info
About Bloomreach
Bloomreach is building the world’s premier agentic platform for personalization.We’re revolutionizing how businesses connect with their customers, building and deploying AI agents to personalize the entire customer journey. And we're building all of that on the intelligence of a single AI engine — Loomi — so that personalization isn't only