LTI’s newly acquired virtual analyst “Leni” could power your analytics adoption

Service provider LTI has announced its acquisition of Lymbyc, an artificial intelligence (AI) product firm based in Bengaluru, India. LTI has made significant investments in acquiring niche digital solutions firms over the last few years, such as Pega consulting firm Ruletronics and analytics startup AugmentIQ, and made the decision to acquire Lymbyc after a year of partnering with the firm on client projects.

Analytics and AI professionals can expect this move to bolster LTI’s intelligent automation platform, Mosaic, with the addition of a virtual assistant “Leni,” Lymbyc’s proprietary AI analyst. HFS sees this acquisition as part of a larger trend to introduce more AutoML (automated machine learning) into AI and automation platforms and environments to make decision sciences more accessible to business. Our research has found that enterprise performance is correlated to those businesses that are making data and insights more accessible to its employees, making this a high priority initiative for business leaders globally.

Lymbyc could make “self-serve analytics” a reality for LTI’s Mosaic client base

Over the last three years, LTI consolidated its IP into Mosaic, a platform that encompasses data engineering, advanced analytics, knowledge-led automation, IoT connectivity, and AI all wrapped up with persona-based experiences for different business and IT users to make sense of enterprise data. For example, Mosaic Catalog and Mosaic Lens are consumption layers that enable less-technical business users to use natural language searching and assisted data wrangling.

LTI had thus far built out various Mosaic components and features based on live client needs. It had also partnered with product specialist ThoughtSpot to integrate search-driven analytics capability into Mosaic to make it easier for business users to query their data with natural language. By going beyond partnering and acquiring Lymbyc, LTI is now putting a stake in the ground and emphasizing Mosaic’s focus on self-serve and assisted analytics and AI—by using AI. Leni, Lymbyc’s virtual analyst, can ostensibly crawl through company data, build hypotheses, apply analytical models, find insights, and present them in easily interpretable stories. Leni features natural language querying, automated alerts to track key business signals and KPIs, and the capability to build out what-if scenario models quickly.

So what does this really mean for current and potential clients of LTI?

If clients buy into the service provider’s Mosaic platform (and accompanied data, analytics, and ML services), they will experience a better way to interact with their data. Mosaic’s capabilities are modular, meaning clients that need help with just certain aspects of their data, analytics, and AI environments can get the benefits without major rip and replacing. The key benefit is that less-technical, business focused users will be able to use Mosaic to do data querying, certain types of analyses, and build their own insight stories for business consumption, instead of relying solely on internal IT or LTI resources.

Note that Leni is positioned as a virtual “analyst” that is specifically geared towards helping users with their data and analytics exploration – it isn’t an all-purpose AI virtual assistant such as IPSoft’s Amelia, or HCLs Lucy.

As an example of AutoML, enterprise clients need to watch out for domain knowledge, adaptability, and explainability

The onward progression of AutoML is inevitable in a market where the technology is complex, the data is siloed and noisy, and the talent is sparse. Tools that can simplify data querying, model building, and visualization are thus gaining significant traction from ML platforms like H2O.ai and RapidMiner to product specialists. The LTI-Lymbyc example fits well within this larger movement.

What clients need to be wary of are one-size-fits-all solutions and relying solely on dumbed-down tools that have a limited set of algorithms, knowledge of domain-specific ontologies, or tuning capabilities. E.g., an AutoML solution that can help select analytical models for standardized use cases will not be able to keep up with new R&D on modeling techniques or specialized use cases. While there are challenges aplenty for business leaders to roll out an AI mandate with scarce talent, the shortcuts mustn’t create an uncontrolled situation or unmanageable beast. In deciding between relying on IT’s or the data science team’s efforts, focus on the problems you are trying to solve and the extent to which they are unique to your business or the extent to which you have unique data to feed the algorithms. Distinguish between those that could be helped by non-technical self-serve modeling and analysis solutions and those that need a more sensitive and hands-on approach.

LTI has been developing a catalog of analytics and ML use cases for specific industry verticals such as oil and gas and functions such as marketing. While Lymbyc maintains that it can “learn any domain based on the best available data,” the company highlights Leni’s applicability specifically in the retail and CPG and life sciences industries. As LTI integrates Leni into the Mosaic ecosystem, it will need to continue to invest in enhancing Leni’s intrinsic knowledge, adaptability, and explainability. Further, any analytics or AI leader that has been on the journey of making their companies more data-driven will bring up the need for change management and process re-imagination when it comes to adoption. LTI will need to match its technology investments with these capabilities to assist clients beyond providing a platform for analytics.

The Bottom Line: LTI’s vision and intent with its purchase of Lymbyc are clear as it tries to make analytics and AI easier for business audiences to consume. Now begins the hard work of execution and effecting change for clients—and that goes beyond technology solutions.

Self-serve analytics tools are certainly getting smarter, and “Leni” is LTI’s latest attempt at bridging the technical skills gap around analytics by using AutoML technology. Evaluate if your use cases can be served by the type of self-assisted analytics that LTI’s Mosaic and Leni have to offer, and if your IT/analytics teams could use the help by letting Leni guide some analyses instead. Ultimately, this type of technology adoption will need to be matched by change management and cultural shifts towards making more informed day-to-day decisions based on data – whether it’s IT or “Leni” doing the assisting.