JOB DESCRIPTION :
Self-Service Analytics (Chat-to-Data / Chat-to-Agent)
Design and build conversational analytics experiences where users ask questions in natural language and receive accurate data, narrative explanations, and visualisations in return.
Develop and tune the semantic and data layer (tables, metrics, relationships, business definitions) so that AI agents can reliably translate natural language into correct queries and outputs.
Build and test chat-to-data flows using tools such as Databricks Genie, Power BI Copilot / Q&A, Copilot Studio, and Azure OpenAI-based agents.
Continuously evaluate and improve query accuracy, response relevance, and visualisation quality returned by the SSA/chat agent (prompt design, grounding data, guardrails, fallback logic).
Define and monitor success metrics for SSA adoption and answer quality (e.g., query success rate, user satisfaction, escalation rate).
Decision Intelligence Workflows
Support the design and implementation of Decision Intelligence workflows across Microsoft Fabric, Databricks & Genie, Power BI, notebooks, Power Automate, Teams, and AI-related tools.
Build and maintain reusable data outputs such as semantic tables, KPI scorecards, gap attribution tables, driver relationship tables, driver trend tables, and alert payloads.
Develop Python / Fabric notebook logic for data preparation, KPI assessment, gap attribution, driver diagnosis, trend monitoring, and AI context generation.
Semantic Model & BI Foundation
Build and maintain Power BI semantic models, including table structure, relationships, DAX measures, refresh logic, KPI definitions, and report/query performance - the same foundation the chat agent relies on to answer correctly.
Prepare clean, reusable, well-documented datasets that can be consumed both by BI reports and by AI prompts, chatbots, or Copilot Studio agents.
Delivery & Collaboration
Connect diagnosis and chat-to-data outputs to business-facing channels: Power BI dashboards, Teams alerts, Outlook messages, chatbot interfaces, or Copilot-based workflows.
Work directly with business stakeholders to translate their decision questions into analytical logic, conversational intents, and structured data outputs.
Document data flows, table definitions, semantic logic, prompt/agent design, and technical dependencies so the SSA solution can be reused and scaled across KPIs, countries, and functions.
Partner with BI
developers, data engineers, analytics translators,
product owners, and business teams to test, evaluate, and continuously improve the chat-to-data experience.
JOB REQUIREMENT
Hands-on experience with Power BI, DAX, Power Query, and semantic model development.
Practical exposure to conversational analytics / chat-to-data tools - e.g., Databricks Genie, Power BI Copilot, Copilot Studio, or similar natural-language-to-query solutions.
Working knowledge of Microsoft Fabric, Lakehouse, Dataflows, notebooks, or similar modern data platforms.
Good Python and SQL skills for data preparation, analysis, and pipeline logic.
Understanding of how semantic layer design affects natural-language query accuracy (naming conventions, metric definitions, relationships, synonyms).
Experience with Power Automate, Teams / Outlook integration, or Microsoft 365 automation is a plus.
Basic to intermediate understanding of AI-enabled workflows, prompt engineering, Copilot, AI Builder, Azure OpenAI, or chatbot/agent concepts is a big plus
Able to structure data outputs for reuse across chat agents, dashboards, and alerts - not only build one-off reports.
Comfortable working with business users to understand performance questions and convert them into data logic and conversational intents.
Good documentation habits and the ability to make technical logic understandable to non-technical users.
Nice to have
Familiar with the FMCG environment, KPI diagnosis, performance management, commercial analytics, RTC, sales execution, revenue management, or supply chain analytics.
Experience creating reusable fact and dimension tables for semantic layers.
Experience preparing structured JSON or tabular outputs for AI prompts or chatbot workflows.
Understanding of data quality checks, refresh monitoring, and basic governance principles.
Exposure to Git, APIs, Azure services, or deployment workflows is a plus.