SITO

AI tender radar for Poland and the EU

SITO for software houses, cybersecurity firms and IT integrators.

Public procurement without noise: filter first, AI second, decision last.

SITO collects notices from TED and Polish sources, normalizes them into one model, rejects weak fits deterministically and only then uses Gemini for score, fit reasons and red flags.

Create account Sign in

Sources

TED + BZP / e-Zamówienia

Order

filter -> shortlist -> AI

For whom

IT, software, cyber, integrators

Decision example

One notice. One recommendation. Clear reasons.

review

Score

86 / 100

Fit

Scope includes SOC, SIEM and 24/7 security monitoring.
Budget sits inside the company profile and does not force hardware-heavy CAPEX.

Red flags

Public sector references are mandatory.
The clarification deadline is short.

Match feed

Live surface

SOC and security monitoring

TED / BZP, deadline, risks and fit reasons

86/100

review

E-services platform for a regional city

TED / BZP, deadline, risks and fit reasons

71/100

watch

Firewall and network infrastructure

TED / BZP, deadline, risks and fit reasons

29/100

skip

This is the AI role in SITO: explain the shortlist, not replace product logic.

Who benefits first

The product is intentionally narrow because broad tender marketplaces generate too much low-value noise.

We start with firms that need to qualify opportunities quickly before investing delivery or presales time.

Software house

System maintenance, platform development, integrations, SLA-heavy work and transformation projects.

Cybersecurity

SOC, SIEM, audits, testing, monitoring, DevSecOps and managed security services.

IT integrator

Mixed-scope projects with service ownership and implementation value, not pure supply-only deals.

How the pipeline works

The logic is simple and defensible: hard criteria first, shortlist second, AI only where the cost-to-signal ratio makes sense.

Collection and normalization

Notices from TED and Polish sources land in one PostgreSQL model, deduplicated by source id and content hash.

Deterministic shortlist

CPV, geography, budget, negative keywords and full-text narrow the market before AI spend happens.

Gemini scoring

Gemini returns a structured result: relevance score, fit reasons, red flags and the recommended next action.

Dashboard and digest

The user gets a match feed, deadlines, sources and a plan-aware email summary.

What you get after entry

The interface should move straight to a decision: what fits, what to skip and what needs attention today.

The first screen combines the company profile, matched tenders, operational signals and plan limits without scattering the workflow across extra modules.

Company profile captured in natural language

Feed with fit, risk and deadline in one row

Sources, queues and digests without a separate ops app

Upgrade point tied to limits and monitoring cadence

Subscription is the core business layer

We are not building pay-as-you-go as the main logic. AI and PDF analysis can extend value, but recurring plans remain the core offer.

Free

1 company profile, limited match list, weekly digest and short history.

Starter

Core

Daily monitoring, full AI explanations, longer history and better shortlist control.

Pro

Multiple profiles, attachment analysis, advanced filters and export for a more mature sales workflow.

MVP scope

SITO is built as a narrow product for daily tender monitoring. The core pieces already work as one operating flow.

Sources and filtering

TED and Polish sources land in one model, and the shortlist is built with hard filters before AI starts working.

Profile and decision

The company profile, fit reasons, red flags and recommended next action are available inside one dashboard.

Monitoring and plan

Digests, subscription, source operations and AI analysis live inside the same product and the same workspace.