Navan Upgrades its AI Chatbot to Target Corporate Travel Cost-Saving

Corporate travel agency Navan aims to expand its market share, using artificial intelligence to its full advantage. 

The company sped out the Generative Predictive Text (GPT) gate in February with the launch of its traveler-facing chatbot Ava. Now, an upgraded version of the automated virtual assistant wants to be the best friend of chief financial officers and corporate finance teams as a real-time data analysis tool able to make cost-saving predictions.

Ava is said to help finance teams using OpenAI’s GPT-4 APIs for admin-facing data analysis, condensing complex real-time spend data and insights to improve corporations’ travel budget bottom lines.

Navan outlined the following examples in the launch release announcing Ava’s new capabilities:

• Analyze travel spending to recommend ways to save money, like leveraging rewards or changing hotel criteria.
•  Summarize travel spend by month and category for distribution with finance teams and chief financial officers.
• Analyze travel spending and compare it to the company’s policy.
• Analyze hotel spending and suggest ways to improve this cost.
• Compare benchmark performance and spending, while maintaining privacy.

Ava is also able to write and send an executive summary of its results — in multiple languages — even a poem if necessary, according to the company.

While Navan, as a tech startup, has been focused on AI technology to enhance its travel management and booking software service, its chatbot development adds an extra shine, and the company doesn’t appear short of investment funds to keep going. 

In October last year, Navan (formerly TripActions) raised $300 million, giving it a valuation of $9.2 billion, according to a company statement at the time. Since then, the corp travel startup has continued to expand through acquisitions, including Spain’s Atlanta Events & Corporate Travel Consultants and its most recent investment into corporate travel startup Tripeur, setting the scene for its market footprint in fast-developing India.  

The overall size of the Navan AI team was not disclosed, but it is led by Chief Technology Officer Ilan Twig, and spans its data, engineering and product departments. The company said it expects to continue leveraging generative AI to enhance Ava’s capabilities but when asked how much it has invested into its Open AI developments thus far, a company spokesperson said, “From the beginning, Navan has leveraged AI across its travel and expense platform to drive automation, predictability, optimization, and world-class customer support for our users — so this is just a continuation of Navan’s regular course of business.”

The Navan spokesperson further stated, “From personalized search results that recognize a user’s travel preferences to automated itemization that instantly categorizes line items on complex receipts, the company’s integration of AI is a key component of its category-defining Business Software Designed for People. Because we have always leveraged AI, integrating GPT is a natural next step in our regular course of action/ approach to software design. “

Ava’s data analysis is accessible through its admin dashboard to select customers in North America and will be rolled out to other markets in the coming months.

“The data insights dashboard with be available to all North America customers in the coming weeks. Prior to rolling out internationally, we are working to ensure our product complies with country-specific regulations,” said a Navan spokesperson. The company did not specify if its acquisitions would benefit from the company’s AI tools and expertise, if at all. 

Whether chief financial officers will value Ava’s capacity to produce data analysis in a poem might seem tone-deaf, it indicates the creativity of customer service chatbots to use large language models (LLMs) to improve conversational search experiences relevant to a company’s customers. The operative word being relevant. A recent AI Impact on Travel report by Skift Research forecasts development and innovation in this area to represent an estimated $1.9 billion market. 

The most powerful (albeit scary) thing about Generative AI and LLMs is their neural-network ability to continuously learn and improve on their outputs and the system efficiencies it is being trained on.  

In the case of corporate travel and expense management, the deeper its access to historical travel purchases and expenditures, the sharper the AI tool’s predictive ability to cut costs and improve budgets. 

Navan is clearly betting on this, giving further tangible return on investment for bean counters who might have doubted the need for travel management software services in the first place. However, employees who might have been abusing or rejigging travel spend to their advantage might soon find themselves more than out of pocket. 

Watch the below video to see how Navan’s admin-facing data analysis chatbot works: