Talroo Energizes Solr Search Engine with Lucidworks

Article Featured Image

Talroo, a job advertising platform provider, helps connect businesses with candidates offering very specific talents, using artificial intelligence to enable its customers to reduce costs per hire.

The 10-year-old company’s success led to fast growth. It was named one of Inc. magazine’s fastest-growing companies in the United States for five consecutive years.

To keep up with growing demand, Talroo wanted to increase its capabilities with Apache Solr, an open-source enterprise search platform, according to Tony Mills, vice president of technology at Talroo, which is based in Austin. “We realized we needed to really make a major jump in how we were going to be utilizing Solr.”

Though the company considered trying to handle the change itself and looked at some other third parties, it kept returning to Lucidworks and its Fusion technology, which it determined would best help it be more efficient and productive with Solr, Mills says. Fusion is part of Lucidworks’ Connected Experience Cloud (CXC) suite of offerings.

Initial discussions with Lucidworks started a couple of years ago, Mills recalls.

“We decided we weren’t going to go off Solr, so we needed something that would make Solr work for us the way we needed it to,” he says. “We decided that we were going to find a partner that would benefit from our unique volume and our data change rate. And we wanted to benefit from a partner’s deployment application lifestyle experience.”

Talroo evaluated Lucidworks’ solution during the last three months of 2019 then signed with the company at the beginning of 2020. The implementation took about three months due to the complexity of switching from an older version of Solr to a newer one.

“We went through some significant system adjustments; we spent a lot of time working with Lucidworks talking through things,” Mills says. “They had a lot of Solr expertise that we could augment to make up for our lack of personnel.”

Lucidworks’ Fusion works as a “wrapper” around Solr, enabling Talroo to make better, more efficient use of the search engine, Mills explains. “Before deploying Lucidworks, it took more than 250 milliseconds to return results to customers, which is relatively slow for our case and required us to pre-scale at scheduled times to handle the fluctuations in request volume. Now with Lucidworks, 95 percent of results are returned under 250 milliseconds across widely varied workloads with no pre-scaling required, which has a direct impact on customer experience. Faster results keep our automated partners happy and ensure our partners come to us first for job postings, which increases our daily revenue.”

With Fusion, Talroo can now consistently manage more than three times the data in the search index from the previous iteration. Lucidworks’ Fusion also provides increased insights into latency issues and increased search engine reliability, functionality, and scale. It lets Talroo use multiple query pipelines and make custom pipelines for different scenarios; take advantage of built-in Solr features and Lucidworks’ machine learning/artificial intelligence during indexing and querying via rewrites and rules and result manipulation; and integrate its own machine learning models directly into the query stream.

Lucidworks’ Fusion has also increased Talroo’s daily revenue by about 10 percent by handling additional search index options not previously available.

Talroo is working with Lucidworks to develop a road map for integrating more of the machine learning tools and historical data to fine-tune recommendations and results from the search engine, Mills says. “There are a lot of things Solr has that we haven’t taken advantage of yet.” 

The Payoff

Since deploying Lucidworks' Fusion, Talroo has seen the following results:

  • increased daily revenue by about 10 percent;
  • consistently managed more than three times the data in the search index; and
  • reduced to below 250 milliseconds the time needed to return 95 percent of search results.

CRM Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues