1Intelligent Data Dictionary Agent 1 PS
Data Dictionary Agent Track
Objective
Build a software-only solution that connects to enterprise databases (Snowflake, PostgreSQL, SQL Server, etc.) and automatically generates comprehensive, AI-enhanced data dictionaries. It should extract schema metadata, analyze data quality, and produce business-friendly documentation. An interactive chat interface should allow users to query and understand the data through natural language.
Challenge
Database documentation is often outdated, incomplete, or non-existent. Technical metadata lacks business context, making it difficult for analysts and business users to understand what data means and how to use it. Your mission is to create a platform that:
- Connects to multiple database types and extracts complete schema metadata (tables, columns, relationships, constraints)
- Performs intelligent data quality analysis with metrics like completeness, freshness, and key health, mathematical statistical analysis of the data on the go would be a plus.
- Uses AI to generate business-friendly summaries and usage recommendations for each table
- Produces documentation in multiple formats (JSON, Markdown) and stores artifacts for future reference
- Provides a conversational chat interface where users can ask natural language questions about their database schema
You may optionally add features to:
- Support incremental updates when schema changes are detected
- Visualize data lineage and table relationships
- Generate SQL query suggestions based on user questions
- Provide data quality alerts and trend monitoring
Tools Allowed
Any software-based stack — no hardware required.
Key Focus
Accurate metadata extraction, meaningful AI-generated business context, and intuitive natural language access to database documentation.
2Business Requirements Document (BRD) Generation Agent 1 PS
BRD Agent Track
Objective
Build a software-only solution that automatically generates comprehensive Business Requirements Documents by ingesting data from multiple communication channels (emails, meeting transcripts, Slack messages, uploaded documents). It should filter noise, extract project-relevant information, and produce a structured, professional BRD. Optional human oversight can be used during review stages. The agent must support iterative editing of generated documents.
Challenge
Business requirements are scattered across emails, meetings, chat messages, and informal documents. Manually synthesizing this information into a coherent BRD is time-consuming and error-prone. Your mission is to create a platform that:
- Integrates with communication tools (Gmail, Slack, meeting transcription services like Fireflies)
- Intelligently filters collected data to extract only project-relevant information (requirements, decisions, stakeholder feedback, timelines)
- Generates a structured BRD with sections including Executive Summary, Business Objectives, Stakeholder Analysis, Functional/Non-Functional Requirements, Assumptions, Success Metrics, and Timeline
- Supports natural language edit requests to modify specific sections or the entire document
- Citation and explainability of the data, through different data structures and algorithms.
You may optionally add features to:
- Detect conflicting requirements across different sources
- Generate requirement traceability matrices
- Summarize stakeholder sentiment and concerns
- Present findings via dashboards or automated status reports
Tools Allowed
Any software-based stack — no hardware required.
Key Focus
Accurate information extraction from diverse sources, intelligent noise filtering, and production of clear, actionable business requirements documentation.
3Data Policy Agent 1 PS
Data Policy Compliance Agent
Objective
Build a software-only solution that ingests free-text PDF policy documents and connects to a company database. It should automatically identify records that violate compliance rules and business policies as outlined in the requirements. Optional human oversight can be incorporated during the data analysis phase. The agent must support periodic monitoring to detect future policy violations.
Challenge
Compliance requirements and business policies are often stored as unstructured documents, while company data changes frequently. Your mission is to create a platform that:
- Ingests and interprets PDF policy documents to extract actionable compliance rules
- Connects to and scans a company database for compliance and business rule violations
- Flags detected violations with clear, explainable justifications
- Incorporates human review and intervention wherever needed
- Periodically monitors data for new or recurring violations
You may optionally add features to:
- Suggest remediation steps for detected violations
- Summarize compliance status and trends over time
- Present findings via dashboards or automated reporting
- Generate audit-ready compliance reports
Tools Allowed
Any software-based stack — no hardware required.
Key Focus
Accurate, explainable, and continuous data policy enforcement with actionable insights.
