TENDER LIFECYCLE · THREE PHASES · THREE PRODUCTS

Find. Evaluate. Award.

The German tender lifecycle has three phases. We built a standalone AI product for each — not an integrated platform. Three products. Three databases. One philosophy: Humans decide. AI structures.

Three phases. Three separate products.

The German tender lifecycle has three phases, and each has its own world: its own tools, its own formats, its own roles. Bidders work with TED alerts and estimation software. Estimators work with PDFs, GAEB files and VOB expertise. Contracting authorities work with VgV, scoring matrices and procurement records.

We deliberately chose against a one-platform-for-everything approach. Instead we build three separate products — auftragr, kalkulr, zuschlagr — each focused on one phase, each with its own database, each usable on its own. Bidders do not need contracting-authority software. Contracting authorities do not see bidder data. The separation is not a compromise — it is a choice.

What holds them together: our team, our procurement expertise, and a shared principle — Humans decide. AI structures.

The shared principle

"AI structures the material. The decision stays with the human. That is not a compromise — that is the design."

Every one of our products produces pre-assessments — not verdicts. Every one documents the source of every recommendation. Every one can be overridden, and the override is logged. Humans decide. AI structures. This principle sits in the footer of every page because it sits in the foundation of every single product.

Infrastructure and model choice

Core infrastructure in the EU

Hosting, database, file storage: Open Telekom Cloud, Frankfurt. Separately for each of the three products. Your data lives on German infrastructure.

Best-in-class language models

For individual analysis steps we deliberately engage the strongest available language model — transparently, with consent, and with pseudonymisation of inputs where appropriate.

GDPR-compliant

EU standard contractual clauses where third-country transfer occurs. No sharing with third parties. No training of AI models on customer data. Deletion on request.

EU AI Act in mind

We design to be compliance-aware — without a marketing certificate. Formal certifications (ISO 27001, BSI C5) are not in current focus. We are open about that.

Frequently asked questions

What is the German tender lifecycle?

The German tender lifecycle has three phases: (1) finding relevant tenders on platforms like TED, Bund.de, DTVP and the state portals. (2) Bidder-side evaluation — go/no-go, risk, calculation. (3) Authority-side — qualification check, price check, scoring and procurement record by the contracting authority. For each of these phases, BlackSwanAI offers its own standalone product.

Is this an integrated platform?

No. We build three separate products — not an integrated platform. Each product has its own database, its own portal, its own purpose. What connects them is not technology but our procurement expertise and the principle "Humans decide. AI structures." We have deliberately chosen against a one-platform-for-everything approach.

Why three separate products instead of one solution?

Because the three phases solve different problems for different audiences. Bidders and contracting authorities have different requirements — legally, procedurally, and in terms of data protection. Separate products stay focused, privacy-friendly and honest. A bidder does not have to use the same software as their contracting authority; a contracting authority does not need access to bidder data. Separate systems keep that clean.

Do the three products share data?

No. The three products have no shared databases. What you set up in auftragr is not inside kalkulr. Bids evaluated in zuschlagr are not visible to auftragr or kalkulr. Each product stands on its own.

Can you still buy the products individually?

Yes — that is the point. Each product can be used on its own. There is no bundle forcing the purchase of all three. Use depends on which phase of the tender lifecycle you work in.

Is this a fully automated procurement process?

No. Every product produces pre-assessments — not verdicts. The human confirms, adjusts or discards. Humans decide. AI structures the material so decisions become faster, more consistent and more traceable.

What infrastructure does BlackSwanAI run on?

Core infrastructure on Open Telekom Cloud (OTC) in the EU. For individual AI analysis steps we deliberately engage the best available language model — with transparent consent, EU standard contractual clauses, and where appropriate, pseudonymisation of input data. Details in privacy policy §7.

Who is behind BlackSwanAI?

BlackSwanAI is developed and operated in Erlangen, Bavaria. Founded by Poonam Nahta and Nitesh Nahta. We are a specialist team for the German tender market.

Start with the product you need

In a 30-minute demo we walk through the 5-lens analysis with kalkulr on a real tender. See how one of our products works in your reality.

Humans decide. AI structures.