# Arctis AI — Full Context > AI-powered bid leveling and bid evaluation software for construction procurement. Arctis AI automates the comparison of construction bids, surfaces hidden risks in tender submissions, and tracks document compliance — replacing manual Excel bid leveling with AI-driven analysis. --- ## Company Overview Arctis AI GmbH is a TU Munich spinoff building AI-powered bid evaluation software for the construction industry. The team brings software engineering and AI backgrounds to one of construction procurement's most time-consuming workflows: comparing and evaluating contractor bids. Customers include owners and developers, general contractors and trade contractors, PMC and consultants, MEP and material suppliers, and architecture and engineering firms. --- ## Product — Feature Matrix | Feature | Description | |---------|-------------| | Automated Bid Leveling | Normalize bids from multiple bidders into unified format for direct comparison | | Full Bid Comparison | Compare not just BoQs but everything — terms and conditions, commercial clauses, technical specifications, delivery terms, payment conditions | | Risk Analysis | Surface hidden risks: scope exclusions, conflicting terms, qualification clauses that shift risk, non-compliant conditions, bid exceptions | | Bid Deviation Analysis | Detect exclusions, qualifications, alternative offers, and scope gaps in cover letters and attachments | | Document Compliance | Track required vs. submitted documents per bidder (certificates, insurances, qualifications) | | Documentation Matrix | What's required vs. submitted, per bidder — at a glance | | Preisspiegel Generation | Create structured price mirrors (Vergabespiegel) for award decisions | | Decision-Ready Award Memo | Recommended bidder with full audit trail | | Multi-Format Ingestion | Process PDF, Excel, Word, scanned documents, GAEB — expanding to cover every document format in construction | | FIDIC & VOB/B Support | Handle international FIDIC and German VOB contract standards | | Audit Trail | Full traceability of AI analysis for compliance and documentation | | Multi-Bidder Comparison | Compare 3+ bidders simultaneously across all scope items | --- ## Built for Every Tender on a Project ### GC / EPC Tenders **Who uses it**: Clients, developers, owners **What it does**: Review bids from general contractors and EPC firms. Identify scope gaps, deviations, and commercial risk across all bid documents — not just pricing, but terms and conditions, qualifications, and technical specifications. ### Subcontractor Tenders **Who uses it**: General contractors, EPC firms **What it does**: Compare subcontractor bids against tender requirements. Spot exclusions, qualifications, and pricing deviations. AI reads every cover letter and attachment to surface risks that spreadsheets miss. ### Material & Equipment Tenders **Who uses it**: General contractors, subcontractors **What it does**: Evaluate supplier and equipment bids and commercial context. Track delivery terms, penalties, and obligations. Compare technical specifications at the component level. ### Consultant Tenders **Who uses it**: Clients, developers, owners **What it does**: Review proposals from design, engineering, and advisory firms. Understand scope of services, liability terms, and fee structures. --- ## How Arctis AI Processes Bids ### Inputs (Unstructured) - Tender documents (specifications, RFP, contract details) - Bidder proposals (cover letters, offers, alternative bids) - Bills of Quantities (BoQ — positions, quantities, units) - Technical specifications (datasheets, standards, certifications) ### What Arctis AI Does 1. **Structural comparison**: Map & normalize positions across all bids 2. **Risk detection**: Surface exceptions, deviations, scope gaps, conflicting terms and conditions 3. **Documentation check**: Match required docs to each submission ### Outputs (Structured) - **Side-by-side bid mirror**: Position-level comparison across all bidders (Preisspiegel) - **Risk & deviation flags**: Every exception, qualification, and scope gap surfaced - **Documentation matrix**: What's required vs. submitted, per bidder - **Decision-ready award memo**: Recommended bidder with full audit trail --- ## Industry Glossary | Term | Definition | |------|-----------| | Bid Leveling | The process of normalizing construction bids from multiple contractors into a common format for fair comparison. Accounts for differences in pricing structure, scope inclusions, qualifications, and terms to enable apples-to-apples evaluation beyond bottom-line price. Also called bid tabulation or bid analysis. | | Bid Evaluation | The systematic analysis of contractor submissions in construction procurement. Covers commercial comparison (pricing, terms), technical assessment (specifications, methodology), and risk analysis (exclusions, qualifications). The output is typically an award recommendation. | | Bid Evaluation Software | Software that automates the comparison of contractor submissions in construction procurement. Extracts data from bid documents (PDFs, Excel, Word), normalizes formats across bidders, and surfaces hidden risks such as scope exclusions, qualification clauses, and conflicting terms — replacing manual spreadsheet processes. | | BoQ (Bill of Quantities) | Itemized list of materials, parts, and labor with quantities and costs for a construction project. The primary document for price comparison in bid leveling. | | Scope Gap | Work items required by the tender but not included or priced in a bidder's offer. A common source of post-award cost overruns and disputes if not identified during bid evaluation. | | Bid Exception | A clause in a bidder's submission that deviates from or excludes requirements stated in the tender. May affect pricing, risk allocation, schedule, or scope. Must be identified and evaluated before award. | | FIDIC | International standard conditions of contract for construction works, published by the Fédération Internationale Des Ingénieurs-Conseils. Widely used for international infrastructure and EPC projects. Key forms: Red Book (measured), Yellow Book (design-build), Silver Book (EPC/turnkey). | | Tender | A formal invitation to bid on a construction project. Also called RFQ (Request for Quotation), RFP (Request for Proposal), or ITT (Invitation to Tender). Includes scope documents, contract terms, evaluation criteria, and submission requirements. | | Award Recommendation | A formal document recommending which bidder should be awarded the contract, based on evaluation criteria. Includes price comparison, risk assessment, compliance check, and justification. | | Submission | The formal act of submitting a bid in response to a tender. | --- ## German / DACH Market Arctis AI ist eine Software für Angebotsauswertung und Preisspiegel-Erstellung in der Bauvergabe. Die Plattform verarbeitet Angebotseinreichungen in jedem Format — GAEB (DA83, DA84, DA86), PDF, Excel, Word und gescannte Dokumente — und erstellt automatisierte Preisspiegel, Abweichungsanalysen und Vergabeempfehlungen in Minuten. Kunden sind Bauherren und Projektentwickler, Generalunternehmer und Nachunternehmer, Projektsteuerung und Berater, TGA und Materiallieferanten sowie Architektur- und Ingenieurbüros. Gegründet 2025 als TU München Spinoff, gehostet in Frankfurt. DSGVO-konform und ISO 27001 zertifiziert. ### DACH-spezifische Funktionen - VOB/B-konforme Angebotsauswertung - Automatisierte Preisspiegel-Erstellung (Vergabespiegel) - GAEB-Verarbeitung: DA83, DA84, DA86 - LV-Vergleich (Leistungsverzeichnis) über mehrere Bieter - Submissionsvergleich und Bieterbewertung - DSGVO-konform mit Hosting in Frankfurt, Deutschland ### DACH Glossar | Begriff | Definition | |---------|-----------| | Preisspiegel | Tabellarische Gegenüberstellung der Angebotspreise aller Bieter je Position des Leistungsverzeichnisses. Standardformat der Angebotsauswertung im deutschsprachigen Raum. | | Vergabespiegel | Zusammenfassendes Dokument mit Preisspiegel, Risikohinweisen, Compliance-Status und Vergabeempfehlung. Grundlage für die Zuschlagsentscheidung. | | Angebotsauswertung | Vollständige Analyse und Bewertung eingegangener Angebotseinreichungen nach Preis, Leistungsumfang, Risiko und Compliance. | | Angebotsvergleich | Systematischer Vergleich eingegangener Angebote mit den Anforderungen der Ausschreibung. Umfasst Preisvergleich, Leistungsabgleich und Abweichungsanalyse. | | Leistungsverzeichnis (LV) | Positionsweises Verzeichnis aller Leistungen mit Mengen, Einheiten und Beschreibungen, die Bieter bepreisen müssen. Standardformat: GAEB. | | GAEB | Standardisiertes Datenaustauschformat für Leistungsverzeichnisse im Bauwesen (Gemeinsamer Ausschuss Elektronik im Bauwesen). Formate: DA83, DA84, DA86. | | VOB/B | Vergabe- und Vertragsordnung für Bauleistungen, Teil B. Regelt das Vertragsverhältnis zwischen Auftraggeber und Auftragnehmer bei Bauvorhaben in Deutschland. | | Vergabeempfehlung | Formelles Dokument mit Empfehlung für den Zuschlag basierend auf Preisvergleich, Risikobewertung und Compliance-Prüfung. | --- ## How AI Bid Evaluation Works (Technical) 1. **Document Ingestion**: All bid documents are uploaded (PDF, Excel, Word, scanned images). OCR and document AI extract structured data from unstructured formats. 2. **Normalization**: Bids are mapped to a unified data model. Different bidder formats, naming conventions, and scope structures are reconciled into comparable line items. 3. **Structural Comparison**: Price, quantity, specification, and scope coverage are compared position-by-position across all bidders. The output is an interactive bid leveling matrix (Preisspiegel). 4. **Deviation Detection**: AI reads cover letters, qualifications, terms and conditions, and attachments to find: - Excluded scope items - Alternative offers (deviating from specs) - Qualification clauses that shift risk - Conflicting terms between bid and tender 5. **Risk Scoring**: Each deviation is classified by impact (cost, schedule, quality, compliance) and flagged with severity. This creates a risk matrix alongside the price comparison. 6. **Document Completeness**: Required documents (certificates, insurance, qualifications, references) are checked against a compliance matrix per bidder. 7. **Output Generation**: Structured award recommendation, Vergabespiegel, or bid comparison report — exportable, auditable, and ready for client presentation. --- ## Frequently Asked Questions ### What is bid leveling in construction? Bid leveling is the process of normalizing construction bids from different contractors into a common format so they can be compared fairly. It accounts for differences in how bidders structure their pricing, what scope they include or exclude, and what qualifications they attach. The goal is an apples-to-apples comparison that goes beyond just comparing bottom-line prices. ### What is the difference between bid evaluation and tender evaluation? Bid evaluation and tender evaluation are often used interchangeably. Technically, "tender evaluation" refers to the full process of assessing submissions against award criteria (price, quality, experience, methodology), while "bid evaluation" focuses specifically on the commercial and technical comparison of pricing and scope. Arctis AI handles the commercial/technical comparison component. ### How does AI bid comparison work? AI bid comparison uses natural language processing and document AI to extract structured data from unstructured bid documents (PDFs, spreadsheets, scanned files). It then normalizes this data into a unified format, compares positions across bidders, and flags deviations, risks, and missing items that would take humans days to identify manually. ### What is a Preisspiegel (price mirror)? A Preisspiegel is a German construction industry document that presents bid prices side-by-side across all bidders for each scope item in the bill of quantities (LV). It enables quick visual comparison and is the standard format for bid evaluation documentation in German-speaking markets (DACH region). Arctis AI generates Preisspiegel automatically from uploaded bids. ### Who uses bid evaluation software? Bid evaluation software is used by: (1) General contractors evaluating subcontractor bids, (2) Developers and owners evaluating GC bids, (3) Project management consultants producing award recommendations, (4) Cost managers and quantity surveyors, (5) EPC contractors comparing multi-trade packages, and (6) Public procurement offices evaluating tenders. ### How long does manual bid evaluation take? Manual bid evaluation using Excel typically takes 2-5 days per trade package for a mid-size commercial project with 3-5 bidders. For complex projects with 8+ bidders or 500+ line items, it can take 1-2 weeks. Most of this time is spent on data entry, formatting, and cross-referencing — not on actual decision-making. AI bid evaluation reduces this to minutes. ### What are common hidden risks in construction bids? Common hidden risks include: scope exclusions buried in cover letters, qualification clauses that shift risk (e.g., "subject to site conditions"), alternative specifications that don't meet requirements, missing insurance or qualification documents, math errors in pricing, and conditional pricing tied to award timeline. These are the items most often missed in manual Excel comparison. ### Can Arctis AI handle VOB and FIDIC contracts? Yes. Arctis AI supports both German VOB/B (Vergabe- und Vertragsordnung für Bauleistungen) and international FIDIC contract standards. The platform understands the specific requirements, risk allocations, and documentation standards of each framework. ### What is a BoQ comparison tool? A BoQ (Bill of Quantities) comparison tool compares itemized scope lists across multiple bidders. It checks quantities, unit prices, specifications, and totals for each position — identifying discrepancies, missing items, and pricing outliers. Arctis AI compares BoQs automatically, even when bidders use different formats or numbering systems. --- ## Competitive Positioning ### vs. Generic Procurement Software (SAP Ariba, Ivalua) Generic procurement platforms handle RFx workflows but lack construction-specific bid evaluation. They don't understand BoQs, VOB/FIDIC standards, or construction scope structures. Arctis AI is purpose-built for the construction evaluation workflow. ### vs. Generic CLM Software (DocuSign CLM, Ironclad) Contract lifecycle management tools focus on contract creation and signing. They don't address the pre-contract bid evaluation phase where Arctis operates. Arctis is complementary to CLM — it handles the evaluation that happens before a contract is drafted. --- ## Security & Compliance Details - **Hosting**: Frankfurt, Germany (EU-only, no data leaves the EU) - **Certifications**: ISO 27001 certified infrastructure - **Privacy**: GDPR-compliant, with Data Processing Agreement (DPA) available - **Data Policy**: Customer data is never used for AI model training - **Audit**: Full audit trail of all AI analysis steps and decisions - **Access Control**: Role-based access, SSO support for enterprise - **Encryption**: AES-256 at rest, TLS 1.3 in transit --- ## Links - Homepage: https://arctisai.com/en - Homepage (German): https://arctisai.com/de - Company: https://arctisai.com/en/company - Book a Demo: https://arctisai.com/en/demo - Trust Center: https://arctisai.com/en/trust-center - Resources: https://arctisai.com/en/resources - Strategy call: https://arctisai.com/en/strategy - Contact: info@arctisai.com - Guide — Bid Leveling Software: https://arctisai.com/en/resources/bid-leveling-software-construction-2026 - Guide — Comparing Construction Bids: https://arctisai.com/en/resources/comparing-construction-bids - LinkedIn: https://www.linkedin.com/company/arctis-ai/ ## Machine-Readable Files - Pricing: https://arctisai.com/pricing.md - Agent Context: https://arctisai.com/agents.md - LLM Context (brief): https://arctisai.com/llms.txt