AI that can be audited
Models work on your documents and data with logged sources, access control and a person who approves the result.
- Home
- Enterprise
Enterprise / AI / Security
Software for enterprise companies: AI, security and NIS2
For companies where software must pass IT, security and procurement reviews: AI in business processes, secure Laravel/PHP applications, integrations with enterprise ERP and support agreements with clear response times.
- Laravel
- PHP 8.3+
- Vue.js / React
- SAML / OIDC
- Docker / CI/CD
Approach
What IT, security and procurement get
Security from day one
Roles, audit logs, dependency scanning and security testing are part of the scope, not an add-on before go-live.
Ready for procurement
NDA, data processing agreement, copyright transfer, architecture documentation and answers to vendor security questionnaires.
Scope
What we do for enterprise teams
AI and LLM in B2B processes
Document search with sources (RAG), data extraction from offers, tenders and invoices, ticket classification and draft answers. We built this in our own product TenderIQ: tender document analysis and an AI bid plan.
Laravel and PHP application security
Security audit against OWASP ASVS, dependency and secret scanning, role and permission review, audit logs, hardening of deployment and coordination with your penetration testers.
NIS2 supply-chain readiness
Under NIS2 essential and important entities must manage the security of their suppliers. We deliver what they need from a software vendor: incident procedures, logs, access control, update policy and documentation.
SSO and enterprise integrations
Single sign-on with Microsoft Entra ID or another SAML/OIDC provider, integrations with SAP, Comarch ERP XL and internal APIs, queues and monitoring.
Support with SLA
Response times agreed per priority, monitoring, backups with restore tests, release windows and runbooks your team can use without us.
Your infrastructure, your code
We deploy to your cloud tenant or on-premise servers, through your CI/CD. Code and documentation belong to you after each stage.
Service scope
How we work with IT, security and procurement
Security review before code
Data classification, threat model and the list of controls are agreed with your security team before development starts.
Environments you control
Separate test and production environments, least-privilege access for our team and access you can revoke at any time.
Documentation for audits
Architecture, data flows, list of personal data processed, dependency inventory (SBOM) and incident procedure.
Clear ownership
A named technical lead, a fixed demo rhythm and decisions recorded in the backlog, so nothing depends on one person’s memory.
Delivery model
How an enterprise project starts
Discovery and security review
1–2 weeks: process, systems, data, risks and the controls the solution must meet.
Proof of value
A narrow scope (for example one AI use case or one integration) on real data, evaluated against agreed criteria.
Delivery in your environment
Staged releases through your CI/CD, with tests, code review and demos every 1–2 weeks.
Handover and support
Documentation, runbooks, training and a support agreement with response times per priority.
Related case study
Open case study
TenderIQ case study: search and AI for tenders
FAQ
Questions before starting
Will our data be sent to AI models outside the EU?
Only if you agree to it. We can use models hosted in the EU, a provider under a data processing agreement or an open-source model on your infrastructure. The choice is made per use case after a data review.
Do you sign an NDA and a data processing agreement?
Yes. NDA, data processing agreement (GDPR art. 28) and copyright transfer to the code are standard parts of our contracts.
Can you help us with NIS2?
We prepare the software side: logs, access control, update and incident procedures and supplier documentation. Legal assessment and the formal audit are done by your compliance team or an external auditor.
Can you deploy in our infrastructure?
Yes. We deploy to your cloud tenant or on-premise servers and work through your CI/CD, repositories and access policies.
What is the smallest sensible first project?
A 1–2 week discovery and security review, or a proof of value for one AI use case or integration. Both end with a written scope, estimate and risk list.
Let’s talk about your project
Start with a 1–2 week discovery and security review. You get a written scope, estimate and risk list.