Future-proof your web forms with Form2Agent AI

What is Form2Agent AI?

Form2Agent AI is a conversational form assistant developed by Freeport Metrics; its purpose is to enable users to complete web forms, carry out queries, and perform data operations using text, voice, and files. It is primarily intended for products and development teams that already have React applications.

Public projects include chat components and multiple sets of presentation forms, but they are not complete hosted SaaS packages. For production use, a backend, model services, business rules, and security mechanisms are still required.

Main functions

Dialog-based form filling

The user answers questions during the chat; the AI extracts the values of relevant fields from the context and inserts them into the corresponding form using client-side scripts. The user can view and modify the results before submitting them.

Voice input and transcription

The assistant can capture audio, convert it to text, and then identify names, dates, amounts, and other fields within that text. Accents, noise, and proper nouns can affect the accuracy.

Voice response

The system can convert the assistant’s responses back into speech, enabling a hands-free operation process. The product page mentions the use of OpenAI TTS-HD; the specific model and associated costs may vary depending on the deployment settings.

Text and file input

In addition to direct chatting, custom implementations can receive files or camera feed to extract structured data from them. The project team is responsible for configuring the settings related to different file formats, sizes, and preprocessing capabilities.

Automatic verification and error correction

Assistants can identify missing fields, ask questions again, correct grammar errors, or rewrite inappropriate text. AI validation can only provide rules based on certainty; it cannot replace backend checks related to types, ranges, and business logic.

Multilingual and real-time translation

The product emphasizes support for multiple languages, accents, and dialects, and it can translate text as well as correct errors in conversations. However, critical contracts, medical, and financial documents still require manual verification or translation by professionals.

Search, filtering, and record updates

Form2Agent allows not only the filling out of new forms, but also the conversion of natural language into actions such as searching, filtering, and updating records. Each action should be accompanied by mechanisms for setting permissions, confirming actions, and allowing them to be reversed.

Can be embedded in existing applications

The public frontend is built using React, which allows chat assistants to be integrated into existing web pages. The development team also needs to create an adaptation layer so that the structured responses correspond accurately to their respective fields and backend interfaces.

Typical inputs and operations

Business scenarios covered in the demonstration

• Healthcare: patient registration, voice assessment reports, and patient search.
• Finance and Insurance: invoices, merchant registration, accident and claim forms.
• Business operations: scheduling demonstrations, creating transactions, and customer filtering.
• Retail e-commerce: Checking orders, updating orders, and adding new products.
• Hotel catering: check-in, kitchen workstations, room inspections, and the front desk.
• Fee process: Travel reimbursement and multi-step applications.

Local installation tutorial

1. Prepare Node.js 21.1.0 or later, as well as npm 10.2.5 or later.
2. Access the public React repository and install the project dependencies.
3. Create a separate OpenAI API key to limit the budget and usage environment.
4. Configure the backend address; by default, the demonstration project uses Freeport Metrics as its backend.
5. Start the frontend and select a form with low sensitivity.
6. Define field names, types, required rules, and client mapping.
7. Test text, voice, empty input, noise, and mid-correction separately.
8. The data is verified repeatedly on the backend, and the user is required to confirm it before it is written.
9. Replace the demonstration backend or establish arrangements for processing production data.
10. Complete tests for permissions, logging, deletion, rate limits, and model costs before going live.

Develop an integrated structure

Prices and fees

A live demonstration is available for direct experience, and the React code can be viewed and downloaded as well. Freeport Metrics has not disclosed the standard subscription price for Form2Agent; custom quotes are provided after contacting their team for commercial integrations.

Free open-source code does not include unlimited model quotas, hosting, maintenance, or compliance support. Costs are also affected by the length of audio conversations, the choice of models, message context, and the backend infrastructure.

Suitable for users and scenarios

• React team: Add conversational interfaces to existing complex forms.
• Accessible products: Offer both text and voice operation methods.
• On-site staff: Enter inspections or reports when it is difficult to operate with the hands.
• Finance and Insurance Team: Assists with multi-step applications and data extraction.
• Hotel retail team: Use dialogues to query or update daily records.
• Prototype team: Use sample code to verify whether the AI user interface is useful.

Advantages

• It is not necessary to completely replace the existing form; it can serve as an additional interaction layer.
• Text, voice, and file input support a wider range of usage scenarios.
• Making React examples public reduces the cost of proof of concept.
• The demonstration scenario shows various actions such as filling in, querying, filtering, and updating.
• Users can still view and modify fields in the traditional interface.

Restrictions and risks

• Public warehouses are primarily used for demonstrating the frontend; they are not complete, self-hosted products.
• By default, the project relies on the Freeport Metrics backend as well as the user’s OpenAI key.
• Speech and models may incorrectly extract amounts, dates, identities, and negative words.
• Automatic field filling by the client cannot replace backend validation and access control.
• Medical, financial, and insurance demonstrations do not equate to ready-made compliance certifications.
• The README explicitly states that sensitive data such as passwords, identification details, or financial information should not be entered.
• The project description states that pull requests and issues will not be reviewed or responded to at the moment.

Data, Privacy, and Security

The README states that the messages sent via the chat interface or the API are saved in order to maintain the context of the current assistant, and it says that this data is not reused in new sessions. It does not specify the exact duration for which the data is stored, the methods used for deleting it, nor the location where the data related to default demonstrations is stored.

The project states that API keys are not stored in the code, but developers still need to take care of protecting those keys themselves. Freeport Metrics’ website privacy policy covers only site contact information and analysis data; it cannot replace the data processing protocols used in production applications.

• Do not submit real sensitive data to public demonstrations.
• In the production environment, a backend proxy is used; model keys are never stored in the frontend.
• Provide clear consent prompts for voice recording and file uploading.
• Set minimum permissions and manual confirmation for each automated action.
• Confirm with the supplier regarding storage, deletion, sub-processors, and data areas.
• Regulated projects must enter into separate data processing agreements and any necessary industry agreements.

Open source and license status

Both the articles on Freeport Metrics and the repository’s README describe Form2Agent as an open-source project under the MIT license; the public repository allows for viewing and forking of the code. Thus, the intention to make this product open source, as well as its open nature, is clear.

However, the root directory of the current repository does not contain a readable LICENSE file, and the code hosting platform has not identified any standard license text. Before copying, modifying, distributing, or using the code for commercial purposes, it is necessary to request the maintainer to provide the actual license text; one cannot rely solely on the README file to determine the rights associated with the code.

API and SDK

Public projects are executed through backend interfaces, but there is no versioned Form2Agent SaaS API or standalone npm SDK available. The packages in the React repository are marked as private, so they cannot be considered published software packages.

When it is necessary to integrate existing systems, one can develop their own adaptation layer based on the examples provided, or contact Freeport Metrics for a customized solution. The interface specifications, authentication methods, data transfer rates, and error handling procedures must all be clearly defined within the project.

Frequently Asked Questions

Is Form2Agent an ordinary chatbot?

It is not just a chat box for answering questions; it analyzes conversations into form fields, search criteria, or record operations, and integrates them with the existing interface.

What inputs are supported?

The current product supports text, voice, and file input; custom implementations can also make use of a camera. The specific format is determined by the deployment settings.

Can npm packages be installed directly?

No public npm SDK has been confirmed; the repository’s package is marked as private. The current approach is to use React example code for integration.

Do you need your own OpenAI key?

Public projects require an OpenAI API key. The actual cost is determined by the model selected and the number of calls made.

Can it handle medical and financial information?

The public demonstration includes relevant scenarios, but the README explicitly warns against entering sensitive data. Production processing requires specialized security and compliance agreements.

Are chat messages saved?

It is saved to maintain the session context; according to the README, it is not reused in new conversations. The exact retention period and method of deletion must be confirmed with the deployment party.

Is Form2Agent open-source from MIT?

The official designation is MIT, but the warehouse currently lacks the actual text of the LICENSE. The scope of legal use can only be determined after obtaining the complete license.

How much is the business version?

There is no standard public price; for customized integrations, it is necessary to contact Freeport Metrics to obtain a proposal and quote.

Can it completely replace traditional forms?

It is not recommended. Keeping visible forms, field validation, and manual confirmation can reduce the risk of identification errors and model mistakes.

Summary

Form2Agent AI adds support for natural language, voice input, and file uploads to existing web forms, making it suitable for rapid verification of interactive data operations. The publicly available React examples serve as a starting point for development, but they are not a complete, ready-to-use production platform.

The key aspects of implementation are backend verification, key protection, session management, and authorization for actions. Before official commercial use, it is also necessary to finalize the license terms and the data processing agreement.