The Architecture That Gets Legal Approval in 17 Days
The problem is not AI. The problem is where the data goes. Every major AI hiring tool sends candidate data to external LLMs for processing. That puts a third-party AI vendor in your supply chain, creates a new attack surface, and triggers 6-12 month legal reviews that usually end in rejection.
Typical AI Hiring Tools
Candidate data sent to vendor cloud for processing
Inference via OpenAI or Anthropic APIs
Third-party AI vendor added to your supply chain
New attack surface from external API calls
6–12 month legal and procurement review
NODES: Built for Your Environment
All data stays inside your VPC or on-prem environment
Open-source models fine-tuned on your data, running on your compute
Zero third-party AI dependencies in your supply chain
No external API calls — no new attack surface
Approved by Fortune 500 legal, CISO, and procurement in production
17 Days
Legal Approval (Fortune 500, after 6 vendors rejected)
34 Days
Contract to Production (last Fortune 500 deployment)
6–12 Months
Typical AI Legal Reviews (industry average)
Book a Demo
NODES consists of six integrated layers. Every layer runs inside your VPC. No layer makes external API calls. The output of each layer feeds the next, creating a closed-loop intelligence system that improves with every hire.
No vendor lock-in. If you leave NODES, you keep everything you built. But ownership is not just about portability — it is about accuracy. Because your models train on your data exclusively, they learn patterns that generic shared models cannot. The intelligence compounds with every hire, and it belongs to you.
Models
You own the trained models, weights, and decision history. If you leave NODES, the models stay with you. Permanently. No re-licensing, no expiration.
Data
All candidate, employee, and performance data stays in your environment. NODES has zero standing access to your production data. Ever.
Encryption Keys
Customer-managed keys via your KMS (AWS KMS, Azure Key Vault, GCP Cloud KMS). NODES never has access to your key material.
Audit Logs
Complete decision history — every score, every decision trace, every bias challenge — stored in your environment. Export to your SIEM anytime.
Infrastructure
Runs on your cloud account or on-prem servers. Your IAM policies, your network rules, your security controls. NODES operates within your security posture, not alongside it.
Connects to Your Existing Stack
ATS Integrations (16+): Ashby, Avature, BambooHR ATS, Factorial ATS, Greenhouse, iCIMS, JazzHR, Jobscore, Jobvite, Lever, Personio ATS, Recruitee, SAP SuccessFactors ATS, Taleo, Teamtailor, Workable, Workday ATS
HRIS / HCM Integrations (28+): ADP Workforce Now, BambooHR, Breathe HR, CharlieHR, Darwinbox, Dayforce, Deel, Deputy, Employment Hero, Factorial, Google Workspace, Gusto, Hibob, HR Partner, HRwork, Humaans, isolved, Keka, MS Entra ID (Azure AD), Nmbrs, Okta, OpportuneHR, Paycom, PeopleStrong, Personio, Rippling, Sage HR, SAP SuccessFactors, SesameHR, Trinet, UKG Pro, UKG Ready, Workday, Xero, Zenefits, Zoho People
SFTP-based data sync available for ADP, isolved, Rippling, and Sage HR where direct API is not available.
DEPLOYMENT OPTIONS
Single-tenant deployment on any infrastructure you control. Same platform, same agents, same DNA engine — regardless of where it runs.
Cloud VPC
AWS, Azure, or GCP. Deploys into your existing cloud account with your security controls, your IAM policies, and your compliance framework.
Government Cloud
AWS GovCloud and Azure Government for FedRAMP, ITAR, and public sector requirements. Same platform, cleared infrastructure.
On-Prem / Air-Gapped
Fully offline deployment for classified environments and organizations that require zero internet connectivity. No external dependencies.
Just as Snowflake defined cloud data warehouses and Databricks defined data intelligence, NODES is defining the Talent Intelligence Infrastructure layer.
Days 1-3
Architecture Review
We map your infrastructure, security requirements, and integration points. You see the full deployment plan before anything is provisioned.
Days 4-10
VPC Deployment.
Platform stack deployed into your environment. Network configuration, security controls, identity provider integration.
Days 11-17
Data Integration
ATS, HRIS, and CRM connectors configured. PII stripping pipeline verified. Data begins flowing into the feature engineering layer.
Days 18-28
DNA Model Training
. Your Top-Performer DNA model is trained on your historical data. Model validation against known outcomes. Bias testing and adversarial review.
Days 29-34
Agent Configuration and Go-Live
Agents configured against your DNA model. Scores write back into your ATS. Your team starts seeing results.
Vendor Access Control
Production Data
NODES cannot access your production data. Not during operation, not during support, not ever. All processing is automated within your environment.
SOC 2 Type II Certified
HIPAA Compliant
Zero-Egress Architecture
Support Access
Customer-initiated only. Time-limited sessions with full audit logging. You approve access, you set the window, you see every action.
Customer-initiated
Time-limited
Fully logged
Updates
NODES provides packaged releases. You deploy to staging first. You test. You approve. You push to production. One-click rollback to any previous version. Your team does not maintain the models — retraining is automated with customer approval gates.
Role-Based Access Control
Customer-Managed Encryption Keys
Encrypted ATS Sync
