Senior Software Engineer-data

Microsoft Microsoft · Big Tech · Hyderabad, TS, IN · Software Engineering

Senior Software Engineer focused on building a petabyte-scale commercial data platform that transforms raw data into intelligence for AI-native agents. The role involves designing and building data pipelines, semantic models, and agent orchestration layers to power decision-making and provide insights through natural language.

What you'd actually do

  1. Design & Build petabyte-scale data pipelines
  2. Convert raw data into intelligence
  3. Design & build agents for leadership and other personas
  4. Deliver KPIs that measure business outcomes
  5. Own engineering fundamentals

Skills

Required

  • Python
  • SQL
  • Spark/Databricks
  • Synapse
  • ADLS Gen2 / Fabric OneLake
  • data modeling
  • semantic data models
  • KPIs
  • data governance
  • security
  • lineage
  • data quality

Nice to have

  • Agent orchestration
  • Copilot Studio Flow
  • FastAPI
  • Azure AI Foundry
  • LLMs
  • SLMs
  • fine-tuned models
  • graph-powered routing and grounding
  • RAG/grounding
  • NL2SQL/NL2DAX
  • skill routing
  • MCP-based integrations
  • evaluation/fine-tuning workflows for agents

What the JD emphasized

  • petabyte-scale commercial data platform
  • design and build the pipelines, curated semantic models, and agents
  • agent orchestration layer
  • AI-native agents
  • 9+ years of software/data engineering experience building production data pipelines at scale
  • Experience building LLM/agentic applications

Other signals

  • design and build the pipelines, curated semantic models, and agents that convert raw data from 200+ transactional sources into trusted KPIs, recommendations, and deep-linked actions across the business
  • work across the full stack of a modern data estate — from ingestion and large-scale compute, to curated logical models, to the agent orchestration layer that answers natural-language questions with facts, reasoning, and recommendations
  • Develop AI-native agents that decompose natural-language questions, route them to the right domain skill, and return insights driven by reasoning, insights, and recommendations