Ajay Pasu­vu­la, Senior Vice Pres­i­dent and Glob­al SAP Ecosys­tem Leader at HCLTech, has over nine years of expe­ri­ence at the lead­ing IT ser­vices com­pa­ny, includ­ing more than three years spear­head­ing SAP ecosys­tem initiatives. 

Under Pasuvula’s lead­er­ship, HCLTech has strength­ened its long-stand­ing part­ner­ship with SAP, marked by ini­tia­tives such as launch­ing an AI Lab in Munich, Ger­many; devel­op­ing AI train­ing pro­grams with SAP; and exe­cut­ing a com­pre­hen­sive plan to cer­ti­fy 15,000 to 20,000 SAP con­sul­tants in gen­er­a­tive AI. These advance­ments posi­tion HCLTech to help enter­pris­es unlock the poten­tial of AI in their SAP environments. 

In an inter­view with ASUG, Pasu­vu­la shared insights on the infu­sion of gen­er­a­tive AI into the SAP ecosys­tem and dis­cussed how HCLTech is help­ing busi­ness­es estab­lish robust data foun­da­tions and opti­mize process archi­tec­tures, enabling them to suc­cess­ful­ly lever­age the full poten­tial of AI with­in the SAP landscape. 

This inter­view, which you can also down­load, has been edit­ed and condensed.

Q: Why do you believe that gen­er­a­tive AI will hold val­ue for IT pro­fes­sion­als? What makes it espe­cial­ly vital for the SAP ecosystem? 

Over the past few years, we’ve seen a mon­u­men­tal shift in the IT indus­try with the rise of gen­er­a­tive AI. It all start­ed when Chat­G­PT became acces­si­ble to end-users with no IT knowl­edge. What makes it dif­fer­ent, from an SAP per­spec­tive, is the abil­i­ty to inte­grate process­es with data and com­bin­ing algo­rithms with large lan­guage mod­els (LLMs). 

SAP’s unique val­ue comes from its vast cus­tomer base of 73,000 cus­tomers and its cen­tral role in glob­al trans­ac­tions. SAP’s posi­tion is unpar­al­leled, with 450,000 cus­tomer data points across SAP S/4HANA, cus­tomer expe­ri­ence, HR, and five decades of process exper­tise. This breadth of data and process knowl­edge cre­ates a sig­nif­i­cant advan­tage from a cus­tomer perspective.

Gen­er­a­tive AI is reshap­ing mod­els based on cus­tomer process­es, data, and best prac­tices as busi­ness mod­els evolve and the indus­try faces dis­rup­tion. SAP’s lead­er­ship is rein­forced by its exten­sive expe­ri­ences across approx­i­mate­ly 23 indus­tries, plac­ing it in a prime posi­tion to offer high­ly rel­e­vant, con­text-dri­ven solutions.

Q: In terms of that spe­cif­ic inte­gra­tion and the rel­e­vance of SAP solu­tions, what prob­lems or hur­dles do you see gen­er­a­tive AI help­ing SAP-equipped enter­pris­es overcome? 

There are mul­ti­ple ways SAP’s land­scape helps cus­tomers trans­form their exist­ing IT and busi­ness process­es. Look at how SAP sup­ports essen­tial back-office func­tions, from order-to-cash, pro­cure-to-pay, and hire-to-retire to com­pre­hen­sive sup­ply chain man­age­ment; most of these process­es rely on SAP data. Then, on the Oper­a­tional Tech­nol­o­gy (OT) side, SAP inte­grates with sys­tems like the man­u­fac­tur­ing cloud and aligns with OT sys­tems; in tele­com, you’ve got the BSS and OSS inte­gra­tion. Giv­en SAP’s piv­otal role across all these indus­tries, gen­er­a­tive AI can sig­nif­i­cant­ly enhance process effi­cien­cy, iden­ti­fy bot­tle­necks, and rec­om­mend trans­for­ma­tive solutions. 

Now, all these LLMs in the mar­ket are pre-trained in indus­try best prac­tices — we call them ver­ti­cal LLMs — and these mod­els must be cus­tomized using the company’s actu­al sys­tem of record data. This is where SAP holds a dis­tinc­tive posi­tion, not just pro­vid­ing best-of-breed process­es from its 50 years of expe­ri­ence but also inte­grat­ing algo­rithms and train­ing mod­els based on real cus­tomer data. That’s huge­ly rel­e­vant from a cus­tomer per­spec­tive, as SAP can think, lis­ten, and act based on their unique data and rec­om­mend new processes. 

For exam­ple, while LLMs excel at cog­ni­tive abil­i­ties like hear­ing, visu­al­iz­ing, and sum­ma­riz­ing, they often lack busi­ness con­text. How does a com­pa­ny apply AI to improve sup­ply chain effi­cien­cy in a customer’s real-world sit­u­a­tion? That’s where SAP dif­fer­en­ti­ates itself. By com­bin­ing LLM capa­bil­i­ties with tab­u­lar func­tion­al­i­ties and a knowl­edge graph, SAP ensures that its solu­tions are not only advanced but high­ly rel­e­vant to each customer’s needs.

Q: What are you see­ing in terms of cus­tomer per­cep­tion of gen­er­a­tive AI? As we look ahead to 2025, are orga­ni­za­tions rec­og­niz­ing its real val­ue or still caught in the hype? 

Many cus­tomers in the man­u­fac­tur­ing and auto­mo­tive sec­tors often ask us how to effec­tive­ly adopt busi­ness AI, espe­cial­ly since SAP is at the core of their dig­i­tal trans­for­ma­tion strat­e­gy. When they approach us, my first ques­tion is, ​“Is your SAP land­scape aligned with your dig­i­tal trans­for­ma­tion goals?” And 99% of cus­tomers say, ​“Absolute­ly. We want SAP to be at the heart of our large dig­i­tal trans­for­ma­tion projects.” 

My sec­ond ques­tion con­cerns how sta­ble their SAP envi­ron­ment is in terms of stan­dard­iza­tion, fit-to-stan­dard, clean core, and inter­face strat­e­gy. Adopt­ing new tech­nolo­gies like AI requires a sol­id infra­struc­ture, includ­ing a uni­fied data mod­el. You also need a clean, lean core that seam­less­ly inte­grates your SAP and non-SAP landscape. 

If you’ve got these three lay­ers struc­tured and scal­able, infus­ing AI into your data makes sense. But if you’ve poor data qual­i­ty and a weak strat­e­gy, even the best LLMs and pre-trained mod­els won’t help. In such a case, you’ll expe­ri­ence hal­lu­ci­na­tions — where the qual­i­ty of out­put will not match the qual­i­ty of the input. This is why SAP’s stan­dard ref­er­ence archi­tec­ture is essen­tial for help­ing cus­tomers infuse AI properly. 

SAP’s AI strat­e­gy is par­tic­u­lar­ly intrigu­ing because it does not lim­it itself to a sin­gle LLM provider, such as Ope­nAI or Lla­ma. Instead, it selects the best LLM for each spe­cif­ic process. For exam­ple, infus­ing AI in SAP Suc­cess­Fac­tors dif­fers com­plete­ly from apply­ing it in the sup­ply chain. This flex­i­bil­i­ty helps cus­tomers extend their out-of-the-box process­es. SAP’s sys­tems even sug­gest which LLM best fits a giv­en process, as the com­pa­ny has rig­or­ous­ly test­ed them all.

Take ABAP chat, for exam­ple. While we call it Chat­G­PT, SAP uses MySQL AI on the back­end for code reme­di­a­tions and func­tion­al­i­ty, much like how HANA GPT is imple­ment­ed. What’s cru­cial here is that SAP con­tin­u­ous­ly refines which LLM is best suit­ed for each spe­cif­ic process. Under­stand­ing this is key for cus­tomers to ful­ly adopt SAP’s offerings. 

Q: Could you talk about what advan­tages the HCLTech and SAP part­ner­ship are bring­ing to clients look­ing to lever­age gen­er­a­tive AI in this area?’

HCLTech and SAP have a long­stand­ing and deeply col­lab­o­ra­tive part­ner­ship. We’re one of the largest Glob­al Strate­gic Ser­vices Part­ner (GSSP) part­ners and the first to embrace RISE with SAP as a ref­er­ence archi­tec­ture. Our entire land­scape runs on SAP, includ­ing Ari­ba, Suc­cess­Fac­tors, Con­cur, Field­glass, RISE with SAP, and BRIM. We’re launch­ing a strate­gic ini­tia­tive on total work­force man­age­ment, where we are build­ing a skill devel­op­ment mod­ule with SAP. With 225,000 employ­ees, we will explore gen­er­a­tive AI use cas­es first­hand for work­force management. 

Our unique posi­tion as both a cus­tomer and a part­ner sets us apart. Based on our expe­ri­ences, we offer gen­uine, real-time feed­back to SAP and oth­er cus­tomers. As a ref­er­ence cus­tomer for RISE with SAP, we can speak to cus­tomers direct­ly and trans­par­ent­ly about what works and what doesn’t, pro­vid­ing unbi­ased insights because we’re active­ly using the solu­tions ourselves. 

At a cor­po­rate lev­el, HCLTech is invest­ing in three key areas: data, AI and SAP. Our CEO announced on Investor Day that SAP will be a sig­nif­i­cant strate­gic focus for us over the next five years. One of our first sig­nif­i­cant invest­ments is the AI Lab in Munich, in col­lab­o­ra­tion with SAP Labs and SAP AppHaus. The lab enables cus­tomers to explore their gen­er­a­tive AI roadmap, engage in design think­ing, build MVPs, and assess the poten­tial for their orga­ni­za­tions. We’re also ensur­ing that all our SAP con­sul­tants — 15,000 to 20,000 pro­fes­sion­als — are cer­ti­fied in SAP gen­er­a­tive AI. We’ve devel­oped a com­pre­hen­sive plan with SAP, focus­ing on indus­tries where we have a strong pres­ence, such as aero­space, util­i­ties, man­u­fac­tur­ing, life sci­ences, and retail. We’re look­ing at each customer’s needs, find­ing gaps in SAP’s roadmap and bring­ing inno­v­a­tive solu­tions to the edges of their busi­ness processes. 

With RISE, the com­ing years are set to be trans­for­ma­tive. We’ve signed a strate­gic Migra­tion Fac­to­ry deal with SAP’s cus­tomer ser­vices team, rec­og­niz­ing the need for 625,000 SAP con­sul­tants for migra­tion over the next three years — an impos­si­ble demand. We’re inte­grat­ing human-bot inter­ac­tions into RISE with SAP method­ol­o­gy to address this. We’re reduc­ing human depen­den­cy while keep­ing humans in the loop, help­ing our cus­tomers accel­er­ate their migra­tions and busi­ness transformations.

Con­nect with Ajay Pasu­vu­la on LinkedIn.

Q: What’s your assess­ment of SAP’s cur­rent gen­er­a­tive AI strat­e­gy? What’s work­ing, what isn’t, and how is HCLTech help­ing to shape their approach?

SAP’s AI strat­e­gy needs clear­er, sim­pler mes­sag­ing to help cus­tomers under­stand its val­ue in a rapid­ly evolv­ing tech land­scape. While SAP high­lights impres­sive fig­ures, such as 23,000 cus­tomer con­sen­sus points and 450,000 plus cus­tomers’ data and process­es, the chal­lenge lies in trans­lat­ing these com­plex capa­bil­i­ties into some­thing tan­gi­ble for cus­tomers. SAP must take this down to the basics, show­ing what infused AI will look like and out­lin­ing a clear, action­able roadmap for the future. 

The gen­er­a­tive AI roadmap can­not be treat­ed like a reg­u­lar prod­uct roadmap. Cus­tomers won’t wait two or three years for AI fea­tures to be enabled in Suc­cess­Fac­tors or Ari­ba. They need AI to move faster and be inte­grat­ed now — not as anoth­er dis­tant fea­ture but as a core com­po­nent of SAP’s offer­ings. The mar­ket is evolv­ing rapid­ly, and SAP must accel­er­ate its AI ini­tia­tives or risk falling behind, espe­cial­ly with the rise of ver­ti­cal solu­tions from hyper­scalers who see data as a uni­ver­sal resource, with SAP as just one piece of the puzzle. 

SAP can’t afford to lose its advan­tage by mov­ing slow­ly. It needs to get AI into its prod­uct roadmap quick­ly, or it will lose the bat­tle to the big­ger hyperscalers. 

How­ev­er, SAP has a dis­tinct advan­tage. They aren’t just rely­ing on LLMs; they’re build­ing their knowl­edge graph and large tab­u­lar mod­els (LTMs) using their vast data. All these invest­ments com­bine unstruc­tured and struc­tured data, mak­ing them unique in their busi­ness context. 

A great exam­ple is the trav­el expense man­age­ment demo, where every­thing flows seam­less­ly from email gen­er­a­tion to book­ing por­tals. With this seam­less, infused AI strat­e­gy, SAP is head­ed in the right direc­tion, but SAP also needs to accel­er­ate its efforts to stay ahead in the com­pet­i­tive AI landscape. 

Q: Can you explain the AI Force Ini­tia­tive and its rel­e­vance to this over­all HCLTech strat­e­gy around AI?

AI Force is how we dri­ve AI inno­va­tion into our cus­tomers’ busi­ness­es, focus­ing on three key areas. The first is ser­vice trans­for­ma­tion, which accounts for 80 – 85% of our busi­ness. This includes ser­vice trans­for­ma­tions and oper­a­tions across var­i­ous regions — SAP S/4HANA migra­tions, cloud migra­tion, appli­ca­tion main­te­nance, engi­neer­ing ser­vices, and infra­struc­ture oper­a­tions. We aim to embed AI into every­thing, mak­ing our con­sul­tants more pro­duc­tive and smart­ly reduc­ing man­u­al tasks. 

The sec­ond area is indus­try-spe­cif­ic, repeat­able solu­tions. With the rise of ver­ti­cal and indus­try-spe­cif­ic LLMs, we’re iden­ti­fy­ing gaps where exist­ing LLMs or even SAP haven’t ful­ly addressed the entire process chain. We’re devel­op­ing solu­tions that can work across mul­ti­ple indus­tries, not just iso­lat­ed cas­es, cre­at­ing scal­able, repeat­able val­ue for our customers. 

The third area is cloud-native appli­ca­tions. We’re inte­grat­ing gen­er­a­tive AI into all our infra­struc­ture ser­vices and cloud-native appli­ca­tions, with sup­port from AI Foundry and Cloud Native Labs. Among all our labs, our SAP AI lab stands out as the only inde­pen­dent-soft­ware-ven­dor (ISV) lab, high­light­ing our strate­gic priorities.

Q: Any part­ing thoughts for the SAP community? 

Before div­ing into new tech­nolo­gies, make sure your data strat­e­gy is sol­id. Data is essen­tial for dri­ving inno­va­tion at the edge. Take RISE with SAP, for exam­ple — it has worked for us because we’ve laid the right foun­da­tions: a strong data strat­e­gy, a robust infra­struc­ture strat­e­gy on Azure, and a clear inte­gra­tion strategy. 

“It’s sim­ple — get your data strat­e­gy right, estab­lish your ref­er­ence archi­tec­ture, and then inte­grate AI. That’s how you achieve mean­ing­ful results.”

With these foun­da­tion­al pieces and a clean core, we can bet­ter pre­dict out­comes when we extend our sys­tems. It’s sim­ple — get your data strat­e­gy right, estab­lish your ref­er­ence archi­tec­ture, and then inte­grate AI. That’s how you achieve mean­ing­ful results.

Vis­it the HCLTech website.