Intel employs over 110,000 people spread across offices and facilities in over 60 countries. With such a global footprint, Intel needs to provide a consistent experience to customers, employees, and partners who speak different languages and come from different cultures - every single day. That's why the company decided to leverage LILT's AI translation solution in order to reduce costs and improve time-to-market.

LILT’s translation services are enabled by the LILT Contextual AI Engine, which uses AI and adaptation to drive productivity and efficiency at every step in the localization process. In only a few months of using LILT, Intel has achieved significant savings on translation costs. LILT’s Contextual AI Engine is instantly retrained using linguist feedback, enabling Intel to translate the same amount of content while reducing costs by 40% year-over-year. The impact of that reduction is significant, enabling one of Intel’s business units to double its volume of translated content with only a marginal increase in budget.

Intel IT had been looking to reinvent the way it approached localization so it could better serve its internal stakeholders. The company had made several attempts to upgrade its localization processes, but those attempts had resulted in incremental - rather than transformational - improvements. When the company came across LILT, it saw potential in LILT’s new way of applying contextual AI to localization.

“We had been exploring our options to support a significant increase in demand for translation and localization services across Intel but wanted to do so at a sustainable cost with a very lean team. It was clear that we needed to be more aggressive in our approach and adopt a more innovative model,” said Loïc Dufresne de Virel, Intel's head of localization.

“When we met the team at LILT, we were encouraged by their commitment to building a better way to localize content by leveraging AI to augment the work performed by human translators. The way we used to operate served us well for many years, but we started to plateau, and we needed to take major steps today to put us ahead of the curve," added Dufresne de Virel.

The team knew that the future of localization meant increasing speed and efficiency and decreasing costs by using AI and automation to reduce or eliminate unnecessary human effort, reserving human skills and expertise for only the highest-value aspects of the process.

Intel’s localization team was frequently being asked to do more with less. As the company continued to grow into new verticals and new locales, the team had to serve new internal business partners with flat or even reduced budgets. With each additional business partner came new subject matter and content types, meaning that the team’s ability to leverage existing linguistic assets like translation memories (TMs) was limited.

In addition, the localization team was not fully leveraging AI and automation. By finding a solution that incorporated potentially transformational technologies like adaptive NMT and intelligent automation, they realized they could get higher ROI on their existing localization spending.

“One of the overarching themes was that my colleagues and I wanted to spend more time thinking about Intel’s localization strategy rather than dealing with the day-to-day overhead and complexities of vendor management. By offering us a very simple and all-inclusive business model, LILT enabled that,” said Dufresne de Virel.

Intel’s goals were to maximize ROI on every element of its localization program - working with a partner who could help drive the fastest translation speeds and the highest quality output - all for the lowest possible cost. On top of that, they wanted a supplier who was investing in R&D to continue to build innovative new functionality in its offering.

Dufresne de Virel’s team consists of localization veterans who have spent decades in the industry and have seen technologies come and go over the years. After considering their options, the team determined that LILT would be best positioned to help them achieve their goals. On top of that, LILT’s vision for the future of localization - as well as its AI-centered strategy for how to create that future - resonated with Intel’s localization team.

Intel conducted an initial proof of concept where it explored LILT's ability to scale rapidly and work with Intel's existing systems. After carefully analyzing the performance and costs/benefits of LILT's solution, they decided to move forward with deployment.

“The productivity gains from LILT’s AI-powered translation services have enabled us to reduce translation costs by 40% year-over-year for the same volume of content - and we expect those cost savings to continue to increase. Because translation makes up a significant share of our overall localization costs, this was very material to the business,” said Dufresne de Virel.

Key to achieving such efficiency gains was the flexibility and extensibility of LILT's Contextual AI Platform and Connectors, which made it easy to connect to Intel’s existing infrastructure and workflows, avoiding the need for expensive custom development.

"LILT’s ability to plug into our existing systems and processes made it possible for us to quickly and easily test their translation services," said Dufresne de Virel.

On top of that, Intel is excited about the possibility to leverage LILT’s automation features to help streamline business processes, reducing the need for human intervention in areas like program management, QA, and review.

LILT has heavily invested in R&D to automate as much manual work as possible across the entire localization value chain. The LILT Contextual AI Engine increases content throughput and reduces the possibility of human error, supported by LILT's innovative contextual AI technology that enables translators to work 3-5x faster with no loss in quality.

"After completing our initial proof of concept in Q3 2019, we ramped up quickly, and the results speak for themselves. In Q1 2020, for one of our largest internal stakeholders, we processed a 2x increase in volume compared to Q1 2019 while keeping our translation costs flat, which is equivalent to a 50% improvement year over year. And while I’m very pleased with such immediate results, I’m more focused about what we can achieve in the longer term as we continue to work with LILT to leverage AI and automation to further improve our localization processes," added Dufresne de Virel.

"[Contextual AI] is the future of translation, as it delivers for most languages an amazing level of fluency and accuracy. Adaptive NMT allows us to address the challenges and limitations of post editing. Each one improves on its predecessor, and our stakeholders are asking us to deploy such solutions at scale," said Dufresne de Virel.

LILT's commitment to industry-leading research in natural language processing and contextual AI helped influence their decision. Intel’s localization team was focused on incorporating contextual AI into its localization process, because more accurate models would result in better localization ROI. But the burden of training and updating language models in order to achieve a high-quality translation was daunting. The consistent performance improvements that LILT makes to its models, however, addressed those concerns.

"We wanted a more automated, more streamlined solution where the [Contextual AI] Engine learns and improves right as we use it. This is what the LILT solution offered us," added Dufresne de Virel.

LILT is laser-focused on driving success for its customers like Intel, keeping an open dialogue about the product vision and roadmap.

“As Intel grows its localization capabilities, we are excited to grow with them. We continuously invest in technology development and are committed to helping customers like Intel maximize the ROI of their localization spend with AI and automation,” said LILT co-founder and CEO Spence Green.

"I run a very lean team and rely heavily on my suppliers’ expertise to constantly innovate and improve. With LILT, I think we found the company that is best positioned to help us leverage the power of [contextual AI]," said Dufresne de Virel.