When we say that the 2017 SAP SAP­PHIRE NOW con­fer­ence was all about the Ital­ian Renais­sance, of course we mean SAP Leonar­do. SAP used the show to launch its ​“sys­tem of inno­va­tion” fea­tur­ing var­i­ous func­tions and competencies.

There’s no acci­dent that SAP chose to name its prod­uct after Leonar­do da Vin­ci. Sig­nore da Vin­ci was known as a poly­math (or per­son of many skills), due to his impres­sive tal­ents that spanned paint­ing, sculp­ture, archi­tec­ture, astron­o­my, math­e­mat­ics, engi­neer­ing, and more.

SAP has used the time in between 2017 SAP­PHIRE NOW and SAP TechEd Barcelona to roll out a series of updates, exten­sions, and enhance­ments to SAP Leonar­do. It’s been work­ing to clar­i­fy how SAP Leonar­do can bring automa­tion effi­cien­cies, intro­duce best prac­tices, deliv­er data effec­tive­ness, and apply all-around appli­ca­tion work­flow excellence.

Get­ting Chat­ty with the Bots

Those of us who thought that SAP might spend more of its time talk­ing about the back-end exten­sions hap­pen­ing in SAP Leonar­do have had our eyes opened. Recent devel­op­ments are still on the back end, but they’re man­i­fest­ing them­selves in front-end functionality.

The com­pa­ny has focused on finess­ing and upgrad­ing the SAP Con­ver­sa­tion­al AI ser­vice inside SAP Leonar­do. This soft­ware com­po­nent is designed to allow cus­tomers to build sophis­ti­cat­ed cor­po­rate chat­bots more eas­i­ly. What is a ​“sophis­ti­cat­ed” cor­po­rate chat­bot? Well it’s not going to offer you pre-din­ner aper­i­tifs or cig­ars after your meal. SAP is sug­gest­ing that its soft­ware can under­stand the busi­ness mod­els that cus­tomers run and can han­dle human con­ver­sa­tions at a deep­er level.

These con­ver­sa­tions can sup­port cus­tomers through a ful­ly auto­mat­ed chat­bot expe­ri­ence that knows the answers to their ques­tions. There also are chat­bot con­ver­sa­tions that can serve inter­nal staff requests for infor­ma­tion — think of this as an auto­mat­ed helpdesk for IT or oth­er departments.

Putting the I in RPA

The type of automa­tion intel­li­gence in SAP Leonar­do runs in line with the firm’s work to invest in Robot­ic Process Automa­tion (RPA). The dif­fer­ence is that SAP likes to call it Intel­li­gent Robot­ic Process Automa­tion (IRPA).

RPA essen­tial­ly exists to auto­mate repet­i­tive busi­ness process­es across a customer’s IT stack using arti­fi­cial intel­li­gence. These sys­tems learn from humans. So, RPA soft­ware will devel­op and refine its exe­cutable actions by ​“watch­ing” a user per­form tasks inside an application’s user inter­face (UI).

The RPA agent is then able to emu­late and repeat those actions. When you plug the RPA into the data ana­lyt­ics mus­cle that you know SAP is pack­ing at the oth­er end of the pipe with SAP HANA, you can see why SAP thinks RPA needs an I to make it more intelligent. 

A Step Toward the Intel­li­gent Enterprise

SAP has also laid down plans to release smart appli­ca­tion design capa­bil­i­ties in the SAP Ana­lyt­ics Cloud solu­tion and has added new part­ner con­tent in the busi­ness con­tent library. ​“With intel­li­gent robot­ic process automa­tion, our cus­tomers will be able to achieve the high automa­tion lev­el nec­es­sary to become an intel­li­gent enter­prise,” said Juer­gen Mueller, chief inno­va­tion offi­cer at SAP.

“Machine learn­ing acts here as the brain that is man­ag­ing excep­tions and guides the RPA bot to exe­cute on desired process­es. With our enter­prise-grade offer for SAP Con­ver­sa­tion­al AI, we have tak­en a step fur­ther than our pre­vi­ous­ly announced indus­try-spe­cif­ic pack­aged bots. The expan­sion allows cus­tomers to auto­mate their cus­tomer sup­port with chat­bots,” Mueller said.

Oth­er Works of (Leonar­do) Art

The last batch of updates were tabled at SAP TechEd Barcelona in Octo­ber 2018. The promise from the SAP Con­ver­sa­tion­al AI team is the abil­i­ty to build fast and high-per­form­ing (and no doubt sophis­ti­cat­ed) cus­tomer sup­port bots in a mat­ter of days.

SAP will now aim to pro­duce ​“brain food” for its bot ven­tures by invest­ing in machine learn­ing (ML) capa­bil­i­ties for data sci­en­tists to train their own cus­tom machine learn­ing models.

“This is part of SAP’s strat­e­gy to move toward a sin­gle solu­tion to deliv­er a uni­fied machine learn­ing expe­ri­ence across the SAP offer­ing, with access to key open source machine learn­ing tech­nolo­gies. This solu­tion is intend­ed to sup­port easy deploy­ment of machine learn­ing mod­els into pro­duc­tion and at scale, eas­ing the bur­den on IT orga­ni­za­tions that man­age the life cycle and inte­gra­tion of machine learn­ing into their cor­po­rate envi­ron­ments,” said SAP in a press statement.

What’s the Big Pic­ture for SAP Leonardo?

Five new machine learn­ing ser­vices are planned for the remain­der of 2018, includ­ing scene text recog­ni­tion and cus­tomiz­able image seg­men­ta­tion, a train­able ser­vice to per­form pix­el-wise clas­si­fi­ca­tion for the detec­tion and recog­ni­tion of objects in pic­tures and their shapes.

As well as work­ing with Google on text-to-speech and speech-to-text ser­vices in this are­na, SAP says it will now extend its con­nec­tion points to the Google Cloud Plat­form through­out the last quar­ter of 2018.

As far as SAP is con­cerned with SAP Leonar­do, the poly­math can keep learn­ing and devel­op­ing new intel­li­gence that is par­tic­u­lar to spe­cif­ic tasks. And also, it is pick­ing up intel­li­gence that is dif­fer­en­ti­at­ed across indus­try ver­ti­cals. Poly­math or not, SAP aims to make SAP Leonar­do a work of art wor­thy of hang­ing in the enter­prise soft­ware gallery of the future.

Curi­ous how you can bring arti­fi­cial intel­li­gence to your cus­tomer sup­port process­es? Watch our record­ed web­casts on SAP Leonar­do. Or read about some real-world exam­ples of machine learn­ing in SAP Leonardo.