Loading...
Loading...
Your scans, BIM models, contracts, and site data are some of the most sensitive assets in your business. Here's exactly how we treat them.
Every retrieval query is scoped to your project. We do not train shared foundation models on your uploads, documents, scans, or conversations. Your data stays yours.
Our retrieval system only ever queries within the project you're working in. There is no cross-project bleed. Permissions are enforced at the project boundary.
Enterprise customers can run the full BETEKK stack inside their own network. Nothing leaves your infrastructure. Ideal for regulated industries and air-gapped environments.
Cloud-hosted workloads run in Singapore (ap-southeast-1). Regional deployments can be arranged for customers with strict residency requirements. SST Terraform-managed infrastructure.
AI suggestions never auto-submit. Every recommendation — defect classification, compliance result, billing code — goes through a human reviewer with explicit approve or reject actions.
Every check run, every compliance result, every agent suggestion is versioned and timestamped. Configurations snapshot at execution time so historical results stay reproducible for audits.
Don't see your question? Ask us directly.
Cloud-hosted workloads run in AWS Singapore (ap-southeast-1). Media files are stored in S3 with optional CloudFront CDN. Enterprise customers can opt for fully on-premise deployment where no data leaves their network at all.
No. We do not train shared foundation models on customer data. Retrieval is project-scoped — your queries never see another customer's documents. If you want to fine-tune a private model on your own data inside your tenant, that's a separate engagement and requires explicit opt-in.
We're aligned to Singapore PDPA by default. SOC 2 Type II and ISO 27001 are on our roadmap — talk to us if these are procurement blockers and we'll accelerate the relevant scope.
Yes. Mutual NDAs and Data Processing Agreements are standard for enterprise engagements. Contact our sales team via the contact page and we'll have paperwork back to you within one business day.
Every AI output — defect classification, compliance check result, billing code suggestion — flows into a review queue with explicit approve/reject actions. Nothing auto-submits. Reviewers are named, timestamped, and the decision is part of the immutable audit trail.
Enterprise procurement teams welcome. We respond to security, privacy, and compliance questions within one business day.