01 · About

About KampAI

KampAI is the brand of UAB Minties kampai — a Lithuanian software company working where software engineering, cyber security and artificial intelligence meet.

demo · example output

$ kampai audit --scope app.example.lt
scanning attack surface ........ done
[ ok ] TLS 1.3 · HSTS · CSP enforced
[ ok ] no secrets in repository
[warn] 2 dependencies with known CVEs
[ ok ] patched · tests passing 42/42
 
$ kampai ship --env production
build signed · deployed · monitoring on
✓ production is live
█

02 · Who we are

Engineers first.

We build software, and we break it — on purpose — so it doesn't break in production. That combination is rarer than it should be: development teams often treat security as someone else's job, and security consultants rarely ship code. We do both.

We work with private companies building new products and with public institutions modernising their systems through public procurement. The approach is the same: understand the problem, build the simplest thing that solves it well, and make it hard to attack.

AI is part of that toolbox, not a slogan. We integrate it where it measurably saves time or improves results, and we treat AI systems with the same security scrutiny as any other software.

The name

„Kampai“ comes from „minties kampai“ — Lithuanian for the corners of thought. The AI in KampAI is not an accident.

03 · Practices

Four practices, one team.

Most problems don't fit neatly into one category. A new product needs security from day one; an AI feature needs solid engineering underneath it. We cover the whole chain so nothing falls between the cracks.

Software development

Web platforms, backends, APIs and internal tools — written to be read, tested and maintained for years, not rewritten in six months.

Web applications & APIs
Backend systems & integrations
Cloud infrastructure & CI/CD
Legacy system modernisation

Cyber security

We look at your systems the way an attacker would, then fix what we find. Security is engineering work, not a checkbox on a form.

Penetration testing & security audits
Secure code review
Infrastructure & cloud hardening
NIS2 & GDPR readiness
Incident response support

AI integration

We connect language models to your data and workflows — safely. Clear evaluation, careful handling of private data, and people stay in control of the decisions that matter.

LLM integration into existing products
Search over internal documents (RAG)
Process automation & AI agents
Security review of AI systems — prompt injection, data leakage

Public procurement

We take part in public procurement (viešieji pirkimai) and deliver IT projects for public institutions, meeting tender requirements on documentation, security and accessibility.

Tenders via CVP IS
Technical proposals & documentation
Security requirements compliance
Long-term support & maintenance

04 · How we work

From first call to production.

01
Scope
We start with the problem, not the technology. Requirements, risks and constraints — on one page.
02
Build
Small, reviewed increments. You see working software early and often.
03
Harden
Threat modelling, testing and review before anything goes live.
04
Operate
Monitoring, updates and support after launch. We stay accountable for what we ship.

05 · Principles

How we think about security.

secure_by_default: Safe defaults, least privilege and a minimal attack surface are design decisions, made at the start — not patches added at the end.
no_black_boxes: We build for handover: code, infrastructure and documentation stay accessible to you. No lock-in.
plain_language: Clear reports with prioritised findings, not 80-page PDFs nobody reads.
your_data: Your data isn't used to train third-party models without your explicit decision.

kampai/policy.yaml

# applied to every project
defaults:
  access: least_privilege
  secrets: never_in_code
  dependencies: pinned_and_audited
  transport: tls_only
  reviews: required
ai:
  human_in_the_loop: true
  train_on_client_data: false
reports:
  format: prioritised_findings