Experience: 5–8 Years
Location: Bangalore
Engagement: C2C | Duration: 3 Months
Budget: ₹2–2.5 LPM
UAN: Mandatory
Role Overview
We are seeking a Data/Applied Scientist (Search) with hands-on experience in Vector Search, LLMs, and advanced search ranking algorithms. The role involves building high-performance ML systems that improve search relevance and retrieval efficiency.
Key Responsibilities
- Develop and deploy ML pipelines for search and ranking applications.
- Work on vector search, hybrid search, and learning-to-rank (LTR) systems.
- Implement embedding generation (BERT, Sentence Transformers, custom models).
- Build and manage embedding indexes using FAISS, ScaNN, or Annoy.
- Work with LLMs for RAG (Retrieval-Augmented Generation).
- Optimize semantic vs lexical search performance.
- Deploy services via Vertex AI, Cloud Run, or Cloud Functions.
- Evaluate models using metrics like Precision@K, Recall, nDCG, MRR.
Required Skills
- Strong in Python, SQL, BigQuery, and PySpark.
- Proficiency with Vertex AI, Matching Engine, Dataproc, ElasticSearch/OpenSearch.
- Solid understanding of Vector Databases and Search Relevance Metrics.
- Experience with CI/CD pipelines and model versioning.
- Familiarity with prompt engineering, embedding optimization, and context windowing for LLMs.
GCP Tools Expertise
- ML & AI: Vertex AI, Matching Engine, AutoML, AI Platform
- Storage: BigQuery, Cloud Storage, Firestore
- Ingestion: Pub/Sub, Cloud Functions, Cloud Run
- Compute: Vertex Pipelines, Dataproc (Spark/PySpark)
- Search: Qdrant, Elasticsearch, OpenSearch
- CI/CD & IaC: GitLab, GitHub Actions
