Built for engineers. Documented for everyone.
AQUNAMA Technical Foundation — Architecture, Security, Deployment
For CTOs, IT leads, and anyone who needs to understand how we build — before we ship.
Our engineering approach
We don't pitch architecture diagrams in a sales meeting and then build something different. We don't rip and replace your existing stack. And we don't sell platforms that lock you in.
Every AQUNAMA deployment follows the same engineering principles:
- checkIntegrate with what you already have — your ERP, CRM, accounting, telephony, document storage
- checkUse modern, enterprise-proven components — no exotic tech we can't support five years from now
- checkSenior engineering and AI-accelerated development — fewer people, faster delivery, higher quality
- checkFixed-scope, fixed-price proposals — you know what you're getting before work begins
- checkCloud-agnostic deployment — AWS, Azure, GCP, on-prem, or our own ISO 27001-certified data center
- checkHonest commercial terms — ownership, access, and licensing agreed clearly upfront, per engagement
The AQUNAMA tech stack
| Layer | What we use |
|---|---|
| Web applications | Next.js (the framework behind Netflix, TikTok, and most modern enterprise web apps) |
| Databases | PostgreSQL (open-source, enterprise-grade) |
| Automation engines | Inngest and similar purpose-built tools for time-delayed, multi-step workflows |
| Authentication | Passwordless (email OTP), SSO, or full custom auth depending on requirements |
| AI / LLM integration | OpenAI, Anthropic (Claude), and other leading model providers |
| Voice AI | Custom voice agents combining STT, TTS, and LLM layers, integrated with telephony providers |
| WhatsApp Business | Via Meta-approved providers (360dialog, Twilio) |
| Email delivery | Enterprise infrastructure (SendGrid, Mailgun, or equivalent) |
| Monitoring | Sentry, Inngest Dashboards, and custom observability layers per project |
Security and compliance
Security is part of the architecture, not a layer added afterward.
- checkSecurity-certified hosting — our own ISO 27001-certified Tier III data center in Prague, plus major cloud provider regions (AWS, Azure, GCP)
- checkData encryption in transit (TLS) and, where applicable, at rest
- checkAccess controls based on the principle of least privilege
- checkMulti-factor authentication for administrative access
- checkAudit logging of access to sensitive data and operations
- checkBackup and disaster recovery with documented RPO/RTO targets per engagement
- checkRegular vulnerability management and dependency monitoring
- checkGDPR-compliant by design — including SCCs for any non-EEA data transfers
We work with clients in regulated industries (insurance, financial services, healthcare) and align deployments with their compliance requirements — including NIS2 readiness, sector-specific frameworks, and internal audit standards.
Data handling and sovereignty
Data sovereignty is a per-project decision. We deploy across EU regions of major cloud providers (AWS, Azure, GCP), our own ISO 27001-certified data center in Prague, or your own infrastructure — based on your compliance, performance, and cost requirements.
Whichever model is chosen, every client gets:
- checkDocumented data flows — you know exactly where your data lives and moves
- checkGDPR-compliant defaults — including data residency, retention controls, and access logging
- checkTransparent sub-processor disclosure
Deployment options — cloud, on-premises, hybrid
Public Cloud (AWS, Azure, GCP, and others)
Deployment on major cloud platforms — typically chosen when clients have existing cloud infrastructure, regional requirements, or want full elasticity.
AQUNAMA Data Center
Hosted in our own Tier III, ISO 27001-certified data center in Prague. Strong fit when clients want EU data sovereignty without managing their own cloud account.
On-premises
For clients with strict regulatory requirements, air-gapped environments, or sovereignty needs beyond the EU. Including GPU-equipped servers where AI inference performance requires it.
Hybrid
Some components on a public cloud or our DC, some on yours. Common in financial services, healthcare, and other regulated industries.
Integration approach
Most AQUNAMA deployments connect to systems you already use:
- checkERPs — SAP, Microsoft Dynamics, Odoo, and custom systems
- checkCRMs — Salesforce, HubSpot, Pipedrive, custom CRMs
- checkAccounting — major Czech and international accounting platforms
- checkTelephony — Twilio, Vonage, Genesys, and PBX systems
- checkDocument storage — SharePoint, Google Drive, Dropbox, custom systems
- checkMessaging — WhatsApp Business API, SMS providers, email infrastructure
- checkCalendar & scheduling — Calendly, Google Calendar, Microsoft 365
The AI layer
We use AI in two distinct ways: inside client deployments and inside our own engineering process.
Inside deployments:
- checkCustom voice agents (STT + LLM + TTS) for inbound and outbound call automation
- checkAI Custom Agents — chatbots, internal assistants, agentic workflows
- checkMultilingual speech analysis for QA and sentiment monitoring
- checkDocument classification, extraction, and processing pipelines
- checkWorkflow orchestration that combines deterministic logic with LLM-driven decisions
Inside our engineering process:
We run an integrated layer of AI agents across Claude, ChatGPT, Hermes, and other platforms to accelerate development, generate test cases, document architecture, and compress timelines.
AQUNAMA's deployment model
| Phase | What happens |
|---|---|
| Discovery | We map your business and find where automation pays back fastest. |
| Scoping | We write a fixed-scope, fixed-price proposal — you know what you're getting before work begins. |
| Build | Senior engineering, AI-accelerated, weekly demonstrable progress. |
| Deploy | We launch with your team — no big-bang surprises, no theoretical rollouts. |
| Maintain | We keep it sharp. Continuous tuning, with commercial and access terms tailored to each engagement. |
Commercial models and ownership
Different projects warrant different commercial models:
- checkCustom one-off builds — typically with full source code in your repository and no platform lock-in
- checkAQUNAMA proprietary platform deployments — where we use our own reusable platform, customized to your operations and hosted by us
- checkHybrid engagements — combining custom development with AQUNAMA components, with clearly defined IP and access boundaries
Want to talk to our engineering team?
Tell us what you're trying to build. We'll bring the right people to the discovery call.
Book a free 30-minute call