Senior Software Engineer - Data Science & AI

Overview

Build reusable AI and Data Science platform components (ranking, retrieval, evaluation, agentic services) for Euromonitor's Passport platform and other customer offerings, as shared services adopted across product teams.

What you'll do

• Develop reusable ranking, personalisation and recommendation components, taking them from experimentation through to shared services with measurable quality baselines.
• Build and operate the evaluation harness as a shared service: golden dataset curation, LLM-as-judge pipelines, regression gates in CI, and quality metrics that consuming teams adopt.
• Advance the shared retrieval capability across structured and unstructured data: chunking and indexing strategies, hybrid search, reranking and embedding lifecycle.
• Bring statistical rigour to the platform: experiment design, judge calibration and the interpretation of quality deltas.
• Design and build backend services and APIs for the AI platform (retrieval, ranking, evaluation and agentic components) as versioned, documented, multi-consumer services.
• Contribute to the agent framework and skills and tool registry, with clean SDK surfaces, backward compatibility and abstraction discipline.
• Build for production from the start: observability, cost and token telemetry, quality gates and Responsible AI controls embedded in every component.
• Onboard consuming teams onto platform components through reference implementations, documentation and paved road patterns.
• Work with Engineering, Architecture and Data Platform teams to keep platform services aligned with Euromonitor's cloud native architecture.
• Raise the bar on engineering practice through design reviews, testing standards and mentoring as the team grows.

What you'll need

• 7+ years across Data Science and software engineering, with deep applied Data Science: Machine Learning fundamentals, ranking, recommendation or NLP experience, and the statistical grounding to design and interpret evaluations.
• A proven track record of building AI and Data Science functionality as reusable services, libraries or components adopted by multiple teams, with evidence of that adoption.
• Hands-on depth in modern Generative AI engineering: RAG and retrieval systems; LLM orchestration frameworks such as LangGraph, LangChain, LlamaIndex or Semantic Kernel; prompt and evaluation tooling; and vector or hybrid search such as Elasticsearch, OpenSearch, Azure AI Search, pgvector or a dedicated vector database.
• Strong backend engineering competence: API and service design, data pipelines, and production operations on a major cloud platform such as Azure, AWS or GCP, with container orchestration on Kubernetes.
• Strong Python, plus working competence in at least one further backend language such as C#, Java, Go or TypeScript.
• Evidence of production rigour: observability and tracing for LLM workloads (OpenTelemetry, LangSmith, Langfuse or similar), automated testing, CI/CD and cost awareness.
• The ability to explain technical approaches, trade-offs, limitations and results clearly to engineers, product owners and other stakeholders.

Nice to have

• Experience with MCP or tool calling protocols, agent runtimes, or LLM gateway patterns such as routing, caching and fallback.
• LLM-as-judge calibration, golden dataset methodology or online experimentation experience, including evaluation frameworks such as Ragas, DeepEval or promptfoo.
• Model development experience with PyTorch, TensorFlow or scikit-learn, including training or fine tuning beyond API level usage of foundation models.
• Exposure to adjacent agent ecosystems such as AutoGen, CrewAI, OpenAI Agents SDK, or MLOps tooling such as MLflow, Azure Machine Learning, SageMaker or Vertex AI.
• Contribution to internal platforms, open source libraries or SDKs with real external consumers.
• Experience in information retrieval or search relevance at scale.
• Background in market research, publishing, information services or another accuracy-sensitive domain.

Details

• Location: Bengaluru, Karnataka, India.
• Work mode: Hybrid.
• Reports to the Technical lead.