5

GovBD-MRA (Market Research Analytics)

GovBD-MRA is an enterprise-grade federal contract intelligence platform that combines Java Spring Boot microservices, Python FastAPI, and Next.js 15 to deliver AI-powered government procurement analytics, entity tracking, and award monitoring for government contractors.

GitHub Website

GovBD-MRA

GovBD-MRA: Federal Contract Intelligence Platform 🏛️

GovBD-MRA (Market Research Analytics) is a sophisticated enterprise platform designed to revolutionize how government contractors discover, analyze, and track federal procurement opportunities. Built as part of the Kontratar ecosystem, MRA combines real-time data ingestion, AI-powered analytics, and intelligent search to provide contractors with actionable insights into government spending patterns, entity relationships, and award trends.

"Transforming government procurement data into strategic business intelligence."

🎯 The Challenge We Solve

Government contractors face significant challenges in the federal marketplace:

  • 📊 Data Overload: 2M+ entities, 1M+ awards annually across multiple sources (SAM.gov, USAspending.gov, FPDS)
  • 🔍 Complex Search: Multi-dimensional queries across entities, awards, agencies, NAICS codes, PSC codes
  • 📈 Analytics Gap: Difficulty identifying spending trends, agency patterns, and competitive landscapes
  • 🤖 Manual Research: Time-consuming manual research and opportunity tracking
  • 🔗 Fragmented Data: Entity data, award data, and historical relationships scattered across systems
  • ⏰ Time Sensitivity: Missing critical opportunities due to delayed notifications
  • 💼 Team Collaboration: Coordinating research across multiple team members and projects

GovBD-MRA solves all of these by providing an integrated, AI-powered platform that automates data aggregation, analysis, and insight generation.

✨ Key Features

🏢 SAM Entity Management

Comprehensive Entity Database

  • 2M+ Registered Entities: Complete SAM.gov entity registry
  • Real-time Synchronization: Daily updates from SAM.gov API
  • Historical Downloads: Automated historical data ingestion
  • Staging Pipeline: Multi-stage validation and deduplication
  • Vector Embeddings: AI-powered semantic search using E5-Large-v2

Entity Data Points

  • Core registration information (UEI, CAGE code, DUNS)
  • Business types and classifications
  • NAICS codes (primary and secondary)
  • PSC codes (Product/Service Codes)
  • Points of contact (POC) with full contact details
  • Physical and mailing addresses
  • Certifications (8(a), HUBZone, WOSB, SDVOSB, etc.)
  • Financial information and banking details
  • Geographical service areas

Advanced Entity Search

  • Full-text search across all entity fields
  • Filter by business type, NAICS, PSC, location
  • Certification filtering (small business, veteran-owned, etc.)
  • Relationship mapping (parent companies, subsidiaries)
  • Export to CSV for bulk analysis
  • Async extract for large datasets

💰 Award Intelligence System

USAspending.gov Integration

  • 1M+ Awards Annually: Complete federal contract awards
  • Real-time ETL Pipeline: Automated daily data extraction
  • Award Staging: Validation and enrichment pipeline
  • Vector Database: Qdrant integration for semantic search
  • Historical Tracking: Multi-year award history

Award Data Coverage

  • Prime contract awards (all federal agencies)
  • Award amounts, dates, and durations
  • Contracting agency and office
  • Recipient information (entity linkage)
  • NAICS and PSC codes
  • Place of performance
  • Contract type and pricing
  • Competition type
  • Set-aside categories

Award Analytics

  • Spending trends by agency, time period
  • Top recipients and contractors
  • NAICS/PSC spending distribution
  • Geographic spending patterns
  • Small business utilization
  • Award size distribution
  • Competition analysis

📊 Advanced Analytics & Drilldown

Automated Analytics Jobs

  • Daily Drilldown: Comprehensive agency spending analysis
  • Weekly Rollups: Historical trend aggregation
  • Scheduled Processing: APScheduler-based automation
  • Incremental Updates: Efficient delta processing

Analytics Dimensions

  • By Agency: Agency-level spending patterns
  • By NAICS: Industry sector analysis
  • By PSC: Product/Service category trends
  • By Entity: Contractor performance tracking
  • By Time: Temporal trend analysis
  • By Geography: Regional spending patterns

Drilldown Capabilities

  • Top 10 recipients per category
  • Spending distribution charts
  • Year-over-year comparisons
  • Award count vs. amount analysis
  • Competition metrics
  • Set-aside utilization

🤖 AI-Powered Chat & Research

Intelligent Conversational Interface

  • LangChain Integration: Multi-LLM support (Ollama, OpenAI, Gemini)
  • Streaming Responses: Server-Sent Events (SSE) for real-time chat
  • Context-Aware: RAG (Retrieval-Augmented Generation) using vector search
  • Memory Management: Automatic conversation history and summarization
  • Multi-Thread: Concurrent conversation threads per user

Research Automation

  • Scheduled Research: Automated research prompt execution
  • Entity Research: Deep-dive entity analysis with AI
  • Award Research: Contract opportunity research
  • Market Analysis: Competitive landscape assessment
  • Trend Identification: AI-powered trend detection

Chat Features

  • Natural language queries across entities and awards
  • File upload support (PDF, DOCX, XLSX)
  • Document Q&A with RAG
  • Export chat history
  • Share research threads
  • Team collaboration on research

🔍 Multi-Source Search Engine

Unified Search Interface

  • Elasticsearch Integration: Fast full-text search across 2M+ records
  • Vector Search: Semantic similarity using Qdrant
  • Hybrid Search: Combined keyword + semantic ranking
  • Fuzzy Matching: Typo-tolerant searches
  • Faceted Filtering: Multi-dimensional filtering

Search Capabilities

  • Entity search (name, UEI, CAGE, DUNS)
  • Award search (title, description, agency)
  • NAICS code search
  • PSC code search
  • Geographic search (state, city, zip)
  • Combined entity + award searches
  • Advanced boolean queries

📁 Project & Document Management

Research Projects

  • Create and organize research projects
  • Associate entities and awards
  • Tag and categorize opportunities
  • Track project status
  • Team collaboration
  • Share project insights

Document Processing

  • Upload RFPs, RFQs, solicitations
  • AI-powered document parsing
  • Extract key information
  • Q&A on uploaded documents
  • Document versioning
  • Attachment management

👥 Team Collaboration & Access Control

Multi-Tenant Architecture

  • Tenant isolation (data segregation)
  • Role-based access control (RBAC)
  • Team management
  • Invitation system
  • Permission management

Team Features

  • Create and manage teams
  • Invite team members
  • Assign roles (admin, member, viewer)
  • Share research and projects
  • Collaborative chat threads
  • Activity tracking

💳 Subscription & Billing

Stripe Integration

  • Multiple pricing tiers
  • Monthly and annual billing
  • Usage-based limits
  • Automatic renewals
  • Payment method management
  • Invoice generation

Subscription Plans

  • Free Tier: Limited searches and entities
  • Professional: Enhanced search, analytics
  • Enterprise: Unlimited access, team features
  • Custom: Tailored solutions for large organizations

🔔 Expiring Opportunities & Alerts

Opportunity Tracking

  • Track expiring solicitations
  • Custom alert thresholds
  • Email notifications
  • Dashboard widgets
  • Favorite opportunities
  • Calendar integration

⭐ Favorites & Watchlists

Personal Tracking

  • Favorite entities
  • Favorite awards
  • Save searches
  • Track competitors
  • Monitor agencies
  • Export watchlists

🏗️ Technical Architecture

GovBD-MRA is built using a modern microservices architecture combining multiple technologies for optimal performance and scalability.

Architecture Overview

The platform utilizes a multi-tier architecture:

  • Frontend Layer: Modern React-based web application
  • Application Layer: Microservices handling business logic and data processing
  • Data Layer: Multiple specialized databases for different use cases
  • External Integration: Connections to government data sources

Technology Stack

Technology Stack

Backend Technologies:

  • Java 17 with Spring Boot
  • Python 3.11+ with FastAPI
  • PostgreSQL for data storage
  • Elasticsearch for full-text search
  • Vector databases for semantic search

Frontend Technologies:

  • Next.js 15 with React 19
  • TypeScript for type safety
  • Modern UI framework with responsive design

AI/ML Technologies:

  • LangChain for LLM orchestration
  • Multiple LLM providers (OpenAI, Google Gemini, Ollama)
  • Advanced embedding models for semantic search
  • RAG (Retrieval-Augmented Generation) pipeline

📊 Data Flow & Integration

The platform integrates data from multiple government sources, processes it through AI-powered analytics, and presents actionable insights to users through an intuitive interface.

Key Data Sources

  • SAM.gov Entity API
  • USAspending.gov Award API
  • FPDS Contract API

Processing Pipeline

  1. Data Ingestion: Automated collection from government APIs
  2. Validation & Storage: Data quality checks and secure storage
  3. AI Enhancement: Vector embeddings and semantic analysis
  4. Search & Discovery: Fast, intelligent search capabilities
  5. Analytics Generation: Automated reporting and insights

📊 Performance & Scale

1. Microservices Architecture

  • 3 Specialized Services: Entity, Award, Backend (chat/analytics)
  • Service Isolation: Each service has dedicated database schemas
  • Independent Scaling: Scale services based on load
  • Technology Diversity: Java, Python, TypeScript in single platform

2. High-Performance Data Ingestion

  • 10 SAM API Keys: Parallel entity fetching (100 req/sec)
  • Batch Processing: 10,000 entities per batch
  • Staging Pipeline: Validate before production insertion
  • Incremental Updates: Only fetch changed entities
  • Historical Downloads: Automated monthly historical data ingestion

3. AI-Powered Intelligence

  • Multi-LLM Support: Ollama (local), OpenAI, Gemini
  • RAG Pipeline: Vector search + LLM generation
  • Intent Classification: Route queries to optimal handlers
  • Streaming Responses: Real-time SSE chat
  • Memory Management: Conversation summarization
  • Web Search: Gemini-powered web search for recent info

4. Vector Search & Embeddings

  • E5-Large-v2: State-of-the-art embedding model (768D)
  • Qdrant Integration: High-performance vector database
  • Semantic Search: Find similar entities/awards by meaning
  • Hybrid Search: Combine keyword + semantic
  • Cosine Similarity: Efficient similarity calculations
  • HNSW Indexing: Fast approximate nearest neighbor search

5. Advanced Analytics

  • Automated Drilldown: Daily and weekly analytics jobs
  • Pre-computed Aggregations: Fast dashboard loading
  • Multi-Dimensional: Agency, NAICS, PSC, Entity, Time, Geography
  • Top-K Analysis: Top 10 recipients per category
  • Trend Detection: Year-over-year comparisons
  • Exportable: Excel export for offline analysis

6. Scalability & Performance

  • Async/Await: Fully asynchronous Python backend
  • Connection Pooling: Database connection reuse
  • Worker Pool: 4 Uvicorn workers in production
  • Compression: Gzip for responses >1KB
  • Caching: TanStack Query caching in frontend
  • Lazy Loading: Infinite scroll for large datasets

7. Enterprise Features

  • Multi-Tenancy: Data isolation per tenant
  • RBAC: Role-based access control
  • Team Collaboration: Shared projects and research
  • Audit Trails: Track all entity/award changes
  • Subscription Management: Stripe integration
  • Usage Limits: Tier-based feature access

📊 Performance & Scale

Data Volume

  • Entities: 2M+ SAM.gov registered entities
  • Awards: 100m+ new awards annually (cumulative 5M+)
  • Analytics: 10K+ pre-computed aggregations
  • Chat Messages: 1M+ AI chat interactions
  • Vector Embeddings: 2M+ entity vectors, 5M+ award vectors

Processing Speed

  • Entity Ingestion: 10,000 entities/min (10 API keys)
  • Award Ingestion: 5,000 awards/min
  • Embedding Generation: 100 embeddings/sec (batch)
  • Vector Search: <50ms for top-10 similarity search
  • Chat Response: 1-2s first token, 50-100 tokens/sec streaming
  • Analytics Drilldown: Complete daily job in 10-15 minutes

API Performance

  • Entity Search: <100ms (indexed queries)
  • Award Search: <150ms (Elasticsearch)
  • Semantic Search: <200ms (vector + rerank)
  • Chat Endpoint: <2s (SSE start)
  • Analytics Dashboard: <500ms (pre-computed data)

Resource Usage

EntityData Service:

  • Memory: 2GB (JVM heap)
  • CPU: 2 cores
  • Database: 50GB (entities + relationships)

AwardLoad Service:

  • Memory: 2GB (JVM heap)
  • CPU: 2 cores
  • Database: 100GB (awards + staging)

MRA Backend:

  • Memory: 1GB (per worker)
  • CPU: 4 cores (4 workers)
  • Database: 20GB (users, projects, analytics)

MRA Frontend:

  • Memory: 512MB (Node process)
  • CPU: 1 core
  • Disk: 500MB (build artifacts)

Databases:

  • PostgreSQL: 200GB total
  • Elasticsearch: 50GB (indexed data)
  • Qdrant: 100GB (vector storage)

🎯 Use Cases

1. Competitive Intelligence

Scenario: Track competitors' contract wins

Workflow:

  1. Search for competitor entities by name
  2. Add to favorites
  3. View award history for each competitor
  4. Analyze spending trends by agency
  5. Identify agencies they frequently win from
  6. Export data for presentation

Result: Understand competitive landscape and target similar opportunities

2. Market Research

Scenario: Identify agencies spending in your NAICS code

Workflow:

  1. Navigate to Analytics dashboard
  2. Filter by NAICS code (e.g., 541512 - Computer Systems Design)
  3. View top agencies by spending
  4. Drill down into agency details
  5. View recent awards in that NAICS
  6. Chat with AI: "What are the trends in DoD IT spending?"

Result: Data-driven agency targeting strategy

3. Opportunity Tracking

Scenario: Monitor expiring solicitations

Workflow:

  1. Set up expiring opportunities alerts
  2. Define threshold (e.g., 7 days before deadline)
  3. Receive email notifications
  4. Review opportunities on dashboard
  5. Add relevant opportunities to projects
  6. Collaborate with team on responses

Result: Never miss critical deadlines

4. Entity Research

Scenario: Deep-dive research on potential teaming partner

Workflow:

  1. Search entity by name/UEI
  2. View entity profile (NAICS, PSC, certifications)
  3. View award history
  4. Create research prompt: "Analyze this entity's past performance"
  5. AI generates comprehensive research report
  6. Share research with team
  7. Export to PDF for meeting

Result: Informed teaming decisions

5. Award Analysis

Scenario: Understand past awards for upcoming RFP

Workflow:

  1. Search for similar past awards by title/description
  2. View award details (amount, dates, awardee)
  3. Identify incumbent contractor
  4. View incumbent's other awards
  5. Chat: "What is the typical contract value for this type of award?"
  6. Export similar awards to Excel

Result: Better pricing and strategy for proposal

🐛 Support

For technical support, feature requests, or bug reports, please contact the Kontratar engineering team.

📝 License

This project is proprietary software owned by Kontratar LLC.

GovBD-MRA - Market Research Analytics Platform
Copyright (C) 2024 Kontratar LLC

All rights reserved. Unauthorized copying, modification, distribution,
or use of this software, via any medium, is strictly prohibited.

🤝 Contributing

GovBD-MRA is a proprietary platform. Contributions are limited to authorized Kontratar team members.

For team members:

  1. Create feature branch: git checkout -b feature/amazing-feature
  2. Make changes and test thoroughly
  3. Run tests: pytest (backend), bun test (frontend)
  4. Update documentation if needed
  5. Commit: git commit -m 'Add amazing feature'
  6. Push: git push origin feature/amazing-feature
  7. Create Pull Request with detailed description
  8. Request code review from team lead
  9. Address review comments
  10. Merge after approval

🛣️ Roadmap

Q2 2026

  • 🔍 Advanced Search: Boolean operators, proximity search
  • 📊 Enhanced Analytics: Predictive spending models
  • 🤖 AI Agents: Autonomous opportunity monitoring agents
  • 📱 Mobile App: React Native iOS/Android apps

Q3 2026

  • 🌐 FPDS Full Integration: Complete contract data from FPDS
  • 📈 Real-time Dashboards: WebSocket-based live analytics
  • 🔗 API Marketplace: Public API for third-party integrations
  • 🎓 Knowledge Base: AI-powered procurement knowledge base

Q4 2026

  • 🧠 Advanced AI: GPT-4-turbo fine-tuned on procurement data
  • 🌍 International Expansion: Support for non-US contracts
  • 🔐 SOC 2 Compliance: Enterprise security certification
  • 📊 Custom Reports: Drag-and-drop report builder

📞 Support

For Issues

  • 🐛 Bug Reports: Contact engineering team
  • 💬 Questions: Internal Slack #mra-support
  • 📧 Email: mra-support@kontratar.com
  • 📚 Documentation: Internal wiki

Service Status

🏆 Project Stats

Codebase Metrics

  • Total Lines: 150,000+ lines
    • Java: 50,000 lines (EntityData + AwardLoad)
    • Python: 40,000 lines (MRA Backend)
    • TypeScript/TSX: 60,000 lines (MRA Frontend)
  • Files: 850+ files
    • Java: 150 files
    • Python: 166 files
    • TypeScript: 450+ files
  • Services: 3 backend services + 1 frontend
  • API Endpoints: 80+ REST endpoints
  • Database Tables: 100+ tables
  • Vector Collections: 3 collections (entities, awards, documents)

Technology Diversity

  • Languages: Java, Python, TypeScript/JavaScript
  • Frameworks: Spring Boot, FastAPI, Next.js
  • Databases: PostgreSQL, Elasticsearch, Qdrant, ChromaDB
  • LLM Providers: Ollama, OpenAI, Gemini
  • Cloud Services: AWS RDS, S3 (planned)

💖 Acknowledgments

GovBD-MRA is built on the shoulders of giants:

  • Spring Team: Excellent enterprise Java framework
  • FastAPI Team: High-performance Python web framework
  • Next.js Team: Revolutionary React framework
  • LangChain Team: LLM orchestration framework
  • Qdrant Team: High-performance vector database
  • Elastic Team: Powerful search engine
  • OpenAI: GPT models powering AI chat
  • Google: Gemini API for web search
  • Ollama: Local LLM runtime
  • SAM.gov: Entity data API
  • USAspending.gov: Award data API

🌟 Why Choose GovBD-MRA?

Comparison with Alternatives

FeatureGovBD-MRAGovWinDeltekBGovSAM.gov
Entity Data✅ 2M+✅ 2M+✅ 2M+⚠️ Limited✅ Native
Award Data✅ 5M+✅ 5M+✅ 5M+✅ 5M+⚠️ Partial
AI Chat✅ RAG-powered❌ No⚠️ Basic❌ No❌ No
Vector Search✅ Semantic❌ No❌ No❌ No❌ No
Analytics✅ Pre-computed⚠️ Basic✅ Advanced✅ Advanced⚠️ Basic
Team Collaboration✅ Full✅ Full✅ Full⚠️ Limited❌ No
API Access✅ REST API⚠️ Paid⚠️ Paid❌ No✅ Free (limited)
Pricing$$$$$$$$$$$$Free (limited)
Self-Hosted✅ Yes❌ No❌ No❌ NoN/A

Ready to revolutionize your government contracting intelligence? Get started with GovBD-MRA today! 🚀🏛️

Built with 💙 by the Kontratar Engineering Team

"Empowering government contractors with data-driven intelligence." ⚡