Best 3rd Party Resume Parser Tool [closed]

Question

We are working on a hiring application and need the ability to easily parse resumes. Before trying to build one, was wondering what resume parsing tools are available out there and what is the best one, in your opinion? We need to be able to parse both Word and TXT files.

Answer 1

I have seen a lot of resumes in PDF format. Are you sure you don't care about them?

I'd recommend something simple:

1. Download google desktop search or similar tool (i.e. Copernic)
2. Drop the files in a directory
3. Point the index tool to that directory, and punch in your search terms.

Answer 2

I suggest looking at some AI tools. Three that I'm aware of are

• ALEX
• Sovren
• Resume Mirror

I think all the products handle Word, txt, and pdf along with a bunch of other document types. Although I've never used it, I've heard unfavorable things about Resume Mirror's accuracy and customer support. I'm a contract recruiter and have used both Sovren's and Hireability's parsers in different ATS's. From my view I thought Hireability did a better job, with Sovren it seemed like I was always fixing errors. And when there was a goof with Hire's I gave it to my ATS vendor and it seemed like it was fixed pretty quickly. Good luck.

Answer 3

Don't try to build one unless you want to dedicate your life to it. Don't re-invent wheels!

We build and sell a recruitment system. I did a long evaluation a few years ago and went for Daxtra - the other one in the frame was Burning Glass but I got the impression that Daxtra did non-US resumes better.

Anyway, we're re-evaluating it. Some parts it does brilliantly (name, address, phone numbers, work history) as long as the resume is culturally OK. But if it's not then it fails. What do I mean: Well, if the resume has as the first line:

Name: Sun Yat Sen

then Daxtra is smart enough to figure out that Sun Yat Sen is the guy's name. (Girl's?)

But if it has as the first line:

Sun Yat Sen

It can't figure it out.

On the other hand if the first line is

Johnny Rotten

then Daxtra works out his name.

Also, it works really well on UK addresses, fairly well on Australian addresses, crashes and burns on Indonesian addresses. That said, we've just parsed 35,000 Indonesian resumes relatively well - CERTAINLY far better than not doing it at all, or doing it manually!

On Skilling: I reckon if someone really tried to make the Skills section work then it would take 3 man-months or so and it would work really well.

Summary: Don't write it yourself, do some really good research on real resumes that you want parsing and dive in.

The key thing is: Don't expect any tool to be anywhere near 100% accurate - but it's a lot better than not having it.

Neil

Answer 4

You may want to have a look at egrabber and rchilli these are two best tools out in the market.

Answer 5

FWIW I just ran 650 international resumes through Rchilli and found the accuracy to be very poor. Names & addresses were mangled and the detail fields were hit and miss.

This was a mix of pdfs & Word docs, primarily from Europe & Asia.

Answer 6

I was wondering if any one update these list. Seems all are 2010 old almost 3 yrs old.

We integrated RChilli, and found them no flaw, support is best, and product is easier to use.

We tested RChilli, Hireability, and Daxtra. Sovren never responded to our emails.

Integration was smooth, and support is best in there.