340+
Monitored sources
14
Specialized agents
9
Enrichment stages
First-Mover
Competitive Advantage
Full-Auto
Fully automated outreach
Customer Background
SAGER AG is a long-standing Swiss manufacturer of thermal and acoustic insulation solutions for the construction industry. The company serves customers throughout Switzerland through production, technical consulting, project support and a nationwide sales organization.
The project was discussed with representatives from sales management, digitalization, finance/CFO, operational management and maintenance. Internal project lead was Marc Loosli, Head of Maintenance, SAGER AG,who coordinated the initiative internally and served as the primary contact between SAGER AG and SERGENTA.
Industry
Insulation Manufacturing
Products
Thermal & Acoustic Insulation
Market
Switzerland-wide
Project Lead
Marc Loosli
Agent Type
Sales Intelligence
Initial Situation
SAGER operates in a market where potential sales opportunities are announced across many fragmented sources. The challenge was not a lack of data — but that the data was distributed, incomplete, unstructured and difficult to compare. A construction publication alone is not a qualified sales lead.
Construction publications
Building permits
Official gazettes
Tenders
Project databases
Commercial register
Company websites
Internal CRM data
Before sales can act, they need to understand:
Solution Overview
The SAGER Sales Intelligence Agent continuously transforms raw external information into structured sales opportunities. A new publication enters the system and is progressively collected, classified, interpreted, enriched, connected with company data, linked with internal knowledge, matched with relevant contacts, qualified, scored and prepared for the sales organization.
The seller doesn't receive a raw publication. They receive a structured opportunity dossier.
Live Dashboard
Echte Screenshots aus dem produktiven SAGER Sales Intelligence Dashboard. Kein Konzept — implementiert und im Einsatz.
Multi-Agent Architecture
14 specialized agents work together — each with a clear task, orchestrated by the Orchestrator Agent.
Orchestrator Agent
The central control layer. Determines which agent acts next, which data sources are queried, whether information is sufficient or further enrichment is needed. Manages workflow state, task dependencies, retries, validation and auditability.
Monitoring Agent
Continuously monitors external sources for new opportunities: construction publications, cantonal and municipal announcements, building applications, building permits, public tenders, official gazettes and structured project feeds.
Publication Intelligence Agent
Interprets the raw publication and structures the information: project name, project type, addresses, client, architect, planner, general contractor, mentioned materials and technical requirements.
Research Agent
Investigates the project and involved organizations further. Searches for official project pages, municipal documents, planning documents, company websites, project announcements and procurement information.
Company Intelligence Agent
Enriches each identified company with structured corporate data: legal name, address, legal form, status, industry, purpose, locations, subsidiaries, group relationships, management and historical changes.
Knowledge Agent
Connects external project data with internal SAGER knowledge: product portfolio, specifications, insulation applications, target customer profiles, reference projects, CRM data, sales regions and internal qualification rules.
Contact Agent
Identifies the most relevant people for the specific opportunity — not generic contact data, but the roles that influence purchasing or specification decisions: client, architect, construction manager, general contractor, facade planner, building physicist.
Relationship Agent
Checks whether SAGER already has a connection to the identified company or person. Analyzes CRM data, previous sales activities, past projects, existing customers, inactive customers, open opportunities and assigned sales representatives.
Product Relevance Agent
Evaluates which SAGER solutions might be relevant for the identified project. Connects project characteristics with building type, construction phase, insulation requirements, thermal performance, acoustics, fire protection and facade or roof applications.
Qualification Agent
Evaluates whether the opportunity is commercially relevant. Considers project type, project size, construction phase, timing, involved companies, product relevance, customer access, decision-maker accessibility, strategic value and estimated revenue potential.
Lead Scoring Agent
Converts the qualification into a prioritized score: commercial potential, product fit, project maturity, timing, relationship strength, contact availability, strategic importance, conversion probability and data confidence.
Sales Strategy Agent
Prepares a recommended approach for the salesperson: who should be contacted first, which project angle is relevant, which SAGER products are suitable, which technical arguments to use and which reference project could support the conversation.
Briefing Agent
Creates the final sales dossier: executive summary, project overview, publication source, involved companies, relevant contacts, company intelligence, existing SAGER relationship, product relevance, opportunity score, reasoning, recommended approach and next action.
Outreach Agent
Prepares — where approved by SAGER — personalized communication: email drafts, call preparation notes, LinkedIn messages, follow-ups and account-specific value propositions. Human-in-the-loop: the salesperson reviews, approves and sends.
Data Enrichment Process
The enrichment process starts with a raw construction publication and ends with a validated, qualified sales opportunity.
Source Data
The system receives basic information: project address, municipality, publication text, applicant, project description and date.
Entity Extraction
Identification of entities: companies, persons, addresses, project names, project roles and technical terms from the publication text.
Entity Resolution
Mapping which legal entity each name corresponds to — distinguishing abbreviations, branch offices, subsidiaries, group companies and similar company names.
Company Enrichment
Enriched company profiles with commercial register and corporate data: legal form, status, industry, purpose, management and group structures.
Contact Enrichment
Identification of relevant decision-makers and project participants with function, role, contact data and decision influence.
Internal Knowledge Enrichment
Matching the project with SAGER products, CRM customers, reference projects, sales regions and strategic priorities.
AI Interpretation
The agents analyze the combined data and assess relevance, opportunity potential, timing, product fit, relationship context and recommended action.
Validation
Checking source reliability, contradictory information, missing fields, confidence thresholds, duplicates and already known projects.
Delivery
The final qualified opportunity is delivered to the responsible salesperson or into the approved sales workflow.
Orchestrator Workflow
The agents do not operate independently. Each step is coordinated and validated by the Orchestrator.
New publication detected
Monitoring Agent captures the entry
Publication Intelligence Agent extracts project data
Orchestrator checks for missing information
Research Agent searches supplementary project information
Company Intelligence Agent enriches companies
Contact Agent identifies decision-makers
Knowledge Agent matches with SAGER knowledge
Relationship Agent checks existing customer relationships
Product Relevance Agent determines product fit
Qualification Agent evaluates the opportunity
Lead Scoring Agent prioritizes
Sales Strategy Agent develops approach strategy
Briefing Agent creates sales dossier
Outreach Agent prepares initial communication
Salesperson reviews and acts
Knowledge Architecture
The system combines three types of knowledge. This combination creates intelligence that no single source could provide alone.
External Knowledge
Internal SAGER Knowledge
Generated Intelligence
Competitive Intelligence Advantage
One of the core values: The platform didn't just collect publications, it built an information advantage over the competition.Every newly published construction project was immediately enriched with intelligence from multiple external and internal sources.
This allowed SAGER to identify relevant stakeholders — clients, architects, planners, general contractors — significantly earlier than traditional sales processes. Sales received an enriched project dossier almost immediately after publication. The result was a valuable first-mover advantage.
Human-in-the-Loop Principle
AI handles
Seller responsible for
Business Value
The central transformation: Sales representatives no longer manually search for fragmented information. They receive enriched, qualified and prioritized opportunities with the relevant companies, contacts, product context and recommended next actions.
Strategic Significance
The SAGER project shows: Enterprise AI is not a single chatbot and not a single autonomous agent. It is an orchestrated system of specialized agents, enterprise knowledge, external data sources, business rules, validation mechanisms, human approval and auditable workflows.
The result is an intelligent sales operating process, specifically engineered for SAGER AG.
Frequently Asked Questions
What is the SAGER Sales Intelligence Agent?
An orchestrated multi-agent system that transforms fragmented construction publication and company data into structured, qualified and prioritized sales opportunities. The salesperson doesn't receive a raw building application, but a complete opportunity dossier with involved companies, relevant contacts, product context and recommended next action.
How does the Orchestrator Agent work?
The Orchestrator is the central control layer. It determines which agent acts next, which data sources are queried, whether the available information is sufficient or whether further enrichment is needed. Without the Orchestrator, the individual agents would operate as isolated tools — it makes them a coherent enterprise workflow.
Which data sources are monitored?
The Monitoring Agent monitors construction publications, cantonal and municipal announcements, building applications, building permits, public tenders, official gazettes and structured project feeds. New entries are automatically captured, classified and deduplicated.
How is the Competitive Intelligence Advantage created?
Instead of just collecting publications, the platform immediately enriches each new construction project with intelligence from multiple external and internal sources. This allows SAGER to identify relevant stakeholders significantly earlier than traditional sales processes. Sales receives an enriched project dossier almost immediately after publication — a valuable first-mover advantage.
What does Human-in-the-Loop mean?
AI handles monitoring, research, structuring, enrichment, prioritization and fully automated, tailored outreach — individually written in SAGER's voice. The salesperson remains responsible for the final commercial decision, relationship management, technical validation and contract closure. The automated initial outreach is written by the Outreach Agent in SAGER style, sent across channels and overridden by the salesperson if needed.
How is it explained how a score came about?
Every assessment must remain explainable. The system always shows why an opportunity received its score — which factors were evaluated, which data sources were used and how high the confidence in the available data is. This is crucial for trust, quality and governance.