Production Environment β Live
Platform Architecture
Enterprise-grade AI infrastructure built on Microsoft Azure. 20+ production services, multi-model failover, SOC 2-ready security posture.
20+
Azure Services
4
AI Models Deployed
3
Failover Tiers
99.9%
Target Uptime
<3s
AI Response Time
System Architecture
Full Stack Overview
Multi-layer architecture from edge delivery to AI intelligence, all running on Azure.
Edge
π Azure Front Door (CDN + DDoS)
π± iOS App (TestFlight)
π» Web App (Next.js 14)
Compute
β‘ App Service P3v3 (32GB / 4 vCPU)
π€ Agents App Service
π― Demo App Service
π Staging Deployment Slot
π¦ Container Registry
AI / ML
π§ Fine-Tuned Cognitive OS
π€ GPT-4o
βοΈ GPT-4o-mini
π text-embedding-3-large
π AI Search (S1)
π Language Service (Text Analytics)
Data
ποΈ PostgreSQL D4ds_v5 (16GB)
β‘ Redis Cache (Standard C1)
π Blob Storage
π Key Vault
π¨ Service Bus
β‘ Event Grid
Networking
π Virtual Network (VNet)
π Private Endpoints (DB)
π Private DNS Zones
Security
π‘οΈ Defender for Cloud
π Microsoft Sentinel (SIEM)
π Microsoft Purview
π Application Insights
π Log Analytics (90-day)
AI Intelligence Layer
Multi-Model Architecture with Automatic Failover
Proprietary fine-tuned model as primary, with cascading fallback to base models. Zero single points of failure.
Request Failover Chain
Tier 1 β Primary
Cognitive OS
Fine-tuned GPT-4.1-mini
~1.5s Β· $0.002/req
~1.5s Β· $0.002/req
β
Tier 2 β Fallback
Current Pipeline
GPT-4o-mini + full prompt
~4s Β· $0.015/req
~4s Β· $0.015/req
β
Tier 3 β Emergency
GPT-4o
Full capability model
~6s Β· $0.035/req
~6s Β· $0.035/req
HNTR Cognitive OS
gpt-4.1-mini (fine-tuned)
Proprietary model trained on 200+ examples of HNTR AI recruiting methodology across 8 categories. Recruiting intelligence embedded in model weights β not prompts. Single-pass JSON generation with deterministic validation.
GPT-4o-mini
gpt-4o-mini-2024-07-18
Cost-efficient base model with full Cognitive OS system prompt. Serves as production fallback with proven pipeline. Handles bulk operations and quick evaluations.
GPT-4o
gpt-4o-2024-11-20
Maximum capability model for complex reasoning tasks. Emergency failover ensures zero downtime even during model deployment updates.
Infrastructure
Azure Service Inventory
20+ production Azure services providing enterprise-grade compute, data, AI, messaging, networking, security, and observability.
β‘ App Service
P3v3
Production web application hosting. 32GB RAM, 4 vCPU. Runs Next.js 14 with App Router, server-side rendering, and API routes. Staging deployment slot for zero-downtime deploys.
Compute
ποΈ PostgreSQL Flexible Server
D4ds_v5
Enterprise relational database. 4 vCores, 16GB RAM, 64GB storage with autogrow. 88 Prisma models across advisors, firms, scoring, pipelines, tenants, agents, communications, and intelligence data.
Data
π§ Azure OpenAI
4 Deployments
Four model deployments: proprietary fine-tuned Cognitive OS, GPT-4o (30K TPM), GPT-4o-mini (150K TPM), and text-embedding-3-large (10K TPM). Multi-tier failover architecture.
AI / ML
π AI Search
Standard S1
Semantic search across advisor profiles, firm intelligence, and behavioral signals. 25GB storage, 36 indexes. Natural language queries powered by embedding vectors.
AI / ML
β‘ Redis Cache
Standard C1
In-memory caching for published firm models, scoring results, pipeline data, and AI response caching. Sub-50ms cache hits eliminate redundant compute.
Data
π Blob Storage
Standard LRS
Object storage for generated playbooks, export archives, roster uploads, and document management. Organized by tenant for multi-tenant isolation.
Storage
π Azure Front Door
Standard
Global CDN with DDoS protection, edge caching, SSL termination, and intelligent routing. Enterprise-grade delivery layer for the web application and APIs.
Compute
π¦ Container Registry
Basic
Private Docker container registry for CI/CD pipeline. Build, store, and deploy container images with automated vulnerability scanning.
Compute
π Key Vault
Standard
Centralized secrets management for API keys, database credentials, encryption keys, and certificates. Hardware-backed security with audit logging.
Security
π‘οΈ Defender for Cloud
All Plans
Unified security management across all resources. Threat detection for servers, databases, storage, containers, AI services, Key Vault, and APIs. Continuous vulnerability assessment.
Security
π Microsoft Sentinel
SIEM
Cloud-native security information and event management. Threat detection, investigation, and automated response across the entire environment. SOC 2 audit trail.
Security
π Microsoft Purview
Governance
Unified data governance and compliance. Sensitive data discovery, classification, and policy enforcement. Regulatory compliance management for financial services.
Security
π Application Insights
Workspace-Based
Real-time application performance monitoring, error tracking, dependency tracing, and custom telemetry. 90-day log retention with Log Analytics integration.
Observability
π¨ Azure Service Bus
Standard
Enterprise message broker for asynchronous communication workflows, event-driven processing, and decoupled service orchestration. Powers the unified communications pipeline and agent task queues.
Messaging
π Language Service
Text Analytics
Azure AI Language for sentiment analysis, key phrase extraction, and entity recognition across advisor communications. NLP-powered intelligence layer for behavioral signal processing.
AI / ML
π Virtual Network
VNet + Private Endpoints
Network isolation with private endpoints for database connectivity. Private DNS zones ensure all data-plane traffic stays within the Azure backbone β no public internet exposure for sensitive resources.
Networking
β‘ Event Grid
System Topic
Event-driven architecture for blob storage triggers, automated document processing workflows, and real-time resource state change notifications across the platform.
Eventing
π€ Agents App Service
Container
Dedicated compute for 21 autonomous business operations agents. Separate deployment from the Recruiter OS, enabling independent scaling and isolated execution for AI-driven workflows.
Compute
π― Demo Environment
App Service
Isolated demonstration environment with synthetic data for prospect presentations and partner evaluations. Separate compute and database from production β zero risk of data cross-contamination.
Compute
π Static Web Apps
Standard
Azure Static Web Apps hosting for the marketing site and application shell. Global edge distribution with automatic SSL, custom domains, and CI/CD integration.
Compute
Security & Compliance
SOC 2 Type II Ready
Enterprise security posture designed for wealth management compliance requirements.
Encryption at Rest & Transit
TLS 1.2+ for all connections. AES-256 encryption for stored data. Azure-managed keys with Key Vault integration.
Entra ID SSO
Enterprise single sign-on with PKCE flow. RBAC with 7 role levels from Platform Admin to Read-Only Viewer.
Threat Detection
Microsoft Defender across all resource types. Sentinel SIEM for automated threat investigation and response.
Data Governance
Microsoft Purview for sensitive data classification, policy enforcement, and regulatory compliance tracking.
Multi-Tenant Isolation
Complete tenant data isolation at the database level. Row-level security policies prevent cross-tenant data access.
Audit Logging
90-day log retention across two Log Analytics workspaces. Full audit trail for compliance reporting and incident investigation.
AI Model Privacy
Fine-tuned model is private to our Azure subscription. Training data never used to improve base models. No data shared with OpenAI.
DDoS Protection
Azure Front Door with built-in DDoS protection, WAF capabilities, and edge-level security filtering.
Proprietary Intelligence
Four Scoring Engines
Each algorithm independently evaluates a different dimension of advisor-firm compatibility and transition probability.
π― HNTR Fitβ’
22 Features
Firm Analyzer algorithm scoring advisor-firm compatibility across production, business model, culture, compensation structure, and growth trajectory. 22-feature weighted scoring system with configurable firm DNA models.
Scoring
π₯ BLIXβ’ Index
Patent Pending
Breakaway Likelihood Index analyzing lifestyle signals, career triggers, compensation events, and sentiment indicators to predict advisor transition probability before competitors detect movement.
Scoring
π‘ HNTR Signalβ’
Patent Pending
Strategic visibility and passive brand familiarity engine. Builds pre-outreach awareness with target advisors through intelligent touchpoint orchestration.
Engagement
β‘ HNTR Watchβ’
Patent Pending
Real-time behavioral monitoring for advisor activity patterns, engagement signals, and trigger events. Automated alerts when timing indicators align for recruiter outreach.
Intelligence
Ready to see it in action?
Schedule a live demonstration of the platform and discuss partnership opportunities.