Structify raises $4.1M seed to turn unstructured web data into enterprise-ready datasets

Brooklinn aimed at one of the most infamous pain points in the world of a basic start, artificial intelligence and information analysts: the process of preparing the pain to prepare data.

Specify Today, it appeared from the stealth mode, and 4.1 million dollars in the people burned $ 4.1 million Bain Capital EnterprisesBy participating in 8vc, Integral Enterprises and strategic angel investors.

The company’s platform uses a visual language model of proprietary called Dora According to industrial surveys, automate the data collection, cleaning and structure, cleaning and structure – a process that usually consumes data scientists to 80%.

“The volume of data available today has exploded completely,” he said in Ronak Gandhi, in the structure, exclusive interview, the founder of the structure. “We hit a large point of infection in the existence of both the blessing and curse. We remain very inaccessible, because we remains unprecedented information, it is very difficult to convert in the correct format for meaningful work decisions.”

Strukify’s approach reflects the width of an industrial industry that represents the solution of information specialists to the “problem of data preparation”. Gartner research indicates that Preparation of insufficient information As one of the main obstacles for the application of a successful AI, it consists of four work with four of the information funds needed to fully capitalize in a generative AI.

How AI-Powered Data Transformation opens in secret work intelligence scale

In Core, structural users allow users to create specific databases by selecting the sources, by selecting the sources and copy this information. Platform, SEC documents and LinkedIn profiles can manage everything from news articles and specialized industrial documents.

According to Gandhi, it was established separately from each other, the model in the house that navigates the Internet as a person.

“This is super high quality. Navigates with things like a human being,” he said. “So we are talking about the quality of human being – the first and most important most important of the principles behind Dora. The Internet reads a person to be a man.”

This approach allows Gandi to support a free perpendic that believes that it will help them to democratize access to structured data.

“Now the way you think about the information, this is a really valuable object,” he said. “I would like to spend so much time and spend a lot of time to walk and walk and walk, and cry, if someone wanted to delete someone.” “

Strukify’s vision to “fund information” – it makes it easy to reset easily if lost.

Up to financial-construction: How are the placement of special databases to solve special problems of industry industry

The company has already taken between many sectors. Using financial groups, construction companies to extract information, converts complex geotechnical documents to read tables and collect real-time organizational graphs for sales groups.

Slater StichThe partner in the capital of the same partner, this versatile partner, the announcement of the financial announcement: “Each company, both are very important and a venture soap, etc.

The diversity of Strukify’s early customer base reflects the universal nature of the problems. According to TechTarget researchData preparation usually covers the series of labor-intensive steps: collection, discovery, profiling, cleaning, structure, transformation and approval – can start before any actual analysis.

Why human experience is important for AI accuracy: “Quadruple Verification” system

The main distinguishes for structuring is the “rectangular inspection” process, which connects the EI with human control. This approach addresses a critical concern in the development of AI: to ensure accuracy.

“When you see something that is suspected of a user or identify some information as a potential suspect, we can send a specialist in case of special use,” he said. “This expert can move in the same way [DoRa]Go to the correct data piece, remove, save, then check if it is correct. “

This process only makes data, as well as training samples that develop the model in domains in specialized as time, especially in domains, especially in domains.

“Those things are so mess.” “I don’t think it will never be a strong understanding of geology in my life. But we are there, and it is a great power and we are able to learn from these experts and put it directly to Dora.”

When data extraction means more powerful, privacy concerns are inevitably emissive. Strukify applied security to eliminate these issues.

“We do not have anything that requires any identification, something that requires an entry, we do not have everything that requires you to go behind some information – our agent is concerned about this privacy,” he said.

The company also prioritizes transparency by providing direct source information. “If you want to learn more about a particular piece of information, you will go directly to this content and see if this black box is in the legacy providers.”

Strukify enters a competitive view of the competitive, which is the result of both the built-in players and other aspects of data preparation. Companies like Alterik, Information, Microsoftand Torture In recent years, there are several specialists, all the information is offered to develop data.

According to CEO Alex Reichenbach, it distinguishes what structurizing, the combination of speed and accuracy. By Reichenbach, they claimed that the last Linkedin post, model optimization and “10x” agent “10X” 10x “by improving the infrastructure.

The company’s start comes in an increasing interest in the EU’s operating automation. According to a TechTarget sheet“Frequent data preparation” is often shown as one of the main investment areas for information and analytical teams, “the capabilities of growing data are increasingly important.

How to apply two friends in practices

For Gandhi, correct the address problems that are first encountered in previous roles.

“The great thing about the structure story of the structure is something both personal and professional,” he said. “I used to say [Alex] When I operate as an information analysis analyst and consultant, this actually spent a lot of time for these workers and their most types of them, companies, company lists and their east coasts, collapsed, itching, data. “

It was especially frustrating that the idea could not quickly interact the database quickly. “What happened to me, you could not sterile and there is some kind of ideas until a quick fashion.”

Its co-founder Alex Reichenbach, was similar to similar problems while working in an investment bank in which the data quality problems are striving to build models on structured information.

What does it plan to use $ 4.1 million for enterprise data preparation

With new financing, it plans to develop the technical staff and build itself as an “information tool in the industry.” The company currently offers both free and paid stations, as well as enterprise options for those who need advanced features such as surface placement or high-specialized data extraction.

As more companies invest in AI initiatives, the importance of high quality, structured data will only increase. Recently MIT TECHNOLOGING SIGHTS REPORT Fourth of five enterprises, the fourth of the information found in the generative AI was not ready to capitalize.

This fundamental problem can open the significant value of this fundamental problem in the industries, for the Gandi and Structuralization team, which solves the problem.

“The fact that you can imagine a world of data sets is an iterative, a kind of foolishness for many users,” he said. “At the end of the day, the pitch consists of this control and customizability.”