Intelligent Automation with AI Agents

The digitalization of public procurement has advanced greatly in the last decade, but digitizing is not always synonymous with automating. Having specifications in PDF instead of paper saves physical space, but it does not save the human time needed to read, analyze, and extract critical data.

Until recently, artificial intelligence was limited to isolated tasks: summarizing a text or answering a question. However, managing a tender is not an isolated task; it is a sequential and complex process. To solve this, GovTech technology has evolved towards Agentic AI.

In this article, we explain what these autonomous agents are, how they differ from a simple chat, and, most importantly, how they orchestrate work to reduce file processing from weeks to hours.

What Is an Autonomous Agent (and Why It Is Not a Chatbot)

To understand the revolution in efficiency, one must distinguish between two technologies:

• The Chatbot (Reactive): It is like an intern to whom you have to give step-by-step instructions. It waits for your prompt, responds, and stops. If the process has ten steps, you need to write ten instructions and supervise each intermediate result.
• The Autonomous Agent (Proactive): It is goal-oriented. You give it a goal ("Analyze the viability of this tender") and the system deploys a series of specialized sub-agents that coordinate with each other, execute sequential tasks, and make logical decisions to complete the work without micromanagement.

How the Tendios Ecosystem of Specialized Agents Works

Tendios does not use a generic model for everything. We employ an ecosystem of autonomous agents where each one assumes a critical function of the file, transforming massive documents into actionable data.

Let's see exactly what each one does:

1. Specifications Agent: Reading and Detection

What does it do? It is responsible for the automatic reading and classification of tender documents (PCAP and PPT). It not only ingests the information but extracts key requirements and detects opportunities relevant to your business, saving you the manual reading of hundreds of pages.

Example of Output (what the agent delivers):

2. Solvency Agent: Capacity Calculation

What does it do? It performs the automatic calculation of financial and technical capacities. It crosses the requirements extracted from the specifications with your company's profile to evaluate the viability of participation in each tender instantly.

Viability Analysis:

• Requirement: Turnover > €300k → Your company: €450k ✔ COMPLIES
• Requirement: ISO 27001 Certificate → Your company: Not available → X CRITICAL ALERT

Agent Recommendation: "This certificate is rectifiable through a commitment to obtain (Art. 76.2 LCSP). Probability of admission: High if commitment is included."

3. Offers Agent: Validation and Proposal

What does it do? It focuses on economic validation and the generation of competitive proposals. It helps optimize prices and offer conditions to maximize your chances of success, aligning the proposal with valuation criteria.

• Action: Generates proposal drafts and validates that economic terms fit within detected competitive margins.

4. Legal Agent: Regulatory Compliance

What does it do? It performs clause verification and regulatory compliance, ensuring legal conformity in all processes. It analyzes the PCAP to detect risks, inconsistencies, or disproportionate requirements.

• Example of detection: "The PCAP requires solvency of 200% of the contract value. This breaches the principle of proportionality. Appeal risk: HIGH."

Practical Case: Orchestrated Workflow (End-to-End)

The true power lies in orchestration. Agents do not work in isolation; they pass the baton to automate the process from start to finish:

1. Detection Phase (Specifications Agent): The system monitors portals and sources, reads documents, and classifies the opportunity as relevant for your company.
2. Viability Analysis (Solvency Agent): Before you decide, the agent has already calculated if you meet the necessary financial and technical capacity.
3. Risk Review (Legal Agent): In parallel, clauses are validated to ensure the specifications comply with regulations.
4. Human Decision: The manager reviews the consolidated report and decides "GO".
5. Proposal Phase (Offers Agent): The agent retrieves data and helps generate a competitive and economically validated proposal.

Result: A process that manually would take days is completed in a continuous flow of minutes.

Use Cases by Organization Type

Automation benefits everyone, but in different ways:

• For SMEs: Allows a single person to manage the volume of tenders that previously required an entire department, moving from analyzing 2 tenders a week to analyzing 10, multiplying their chances of winning.
• For Large Enterprises: Eliminates information silos. The Legal Agent alerts the Offers Agent of a risk, ensuring no one works on an offer that will later be excluded due to a formal defect.
• For Public Administration: Allows evaluating hundreds of offers in record time, using agents to automatically verify if bidders meet specifications, reducing the collapse of contracting committees.

Conclusion: Technology That Works, Not Just Chats

The future of technology in the public sector is not about having a smarter chat, but about having more autonomous processes. Agentic Workflows represent the definitive step to free human talent from repetitive bureaucracy.

Don't look for a tool that only talks to you; look for an infrastructure that works with you.

Do you want to see these agents working in real-time on your files?

If you are a public institution:

If you are a private company: