Down­load the full inter­view here. 

“For­ward-look­ing util­i­ties under­stand that mod­ern­iza­tion can­not come at the expense of the com­mu­ni­ties they serve,” said Deep­ak Garg, Chair­man, Founder, and Group CEO at SEW​.AI.

The ver­ti­cal AI com­pa­ny sup­ports ener­gy and water com­pa­nies with a cloud-native foun­da­tion, pro­vid­ing con­text-spe­cif­ic solu­tions to sup­port the unique grid needs that util­i­ties rely on. Garg leads the cre­ation of indus­try-spe­cif­ic agen­tic AI plat­forms, aim­ing to both effi­cient­ly assist util­i­ties providers and the demands they face, while advo­cat­ing for glob­al sustainability.

In a con­ver­sa­tion with ASUG, Garg explained how the SEW​.AI Plat­form works with SAP to dri­ve fur­ther data-informed insights. He also addressed the numer­ous pres­sures of the util­i­ties indus­try — and how AI can help bet­ter sup­port the com­mu­ni­ties these com­pa­nies serve.

This inter­view has been edit­ed and con­densed for length and clarity.

Q: Many com­pa­nies run SAP as the back­bone for ERP, billing, and asset man­age­ment. How does SEW’s plat­form sit along­side that infra­struc­ture, and what does the inte­gra­tion look like?

Our part­ner­ship with SAP spans decades; it’s a part­ner­ship we are proud of and one that has been built on a deep align­ment in how we view this indus­try. SAP has con­sis­tent­ly been the back­bone for ener­gy and util­i­ties, bring­ing struc­ture, reli­a­bil­i­ty, and scale to some of the most com­plex oper­a­tions in the world. As the indus­try has evolved, that foun­da­tion has only become more valu­able. At the same time, what has changed is the envi­ron­ment around it. Util­i­ties today are oper­at­ing in a far more dynam­ic, data-rich, and unpre­dictable land­scape. The oppor­tu­ni­ty we saw was not to rethink that foun­da­tion, but to extend its value.

SEW​.AI Plat­form works along­side as extend­ed SAP plat­form, and is deeply inte­grat­ed into its data and process­es, bring­ing addi­tion­al lay­ers of intel­li­gence across the ener­gy and util­i­ty val­ue chain to enable faster, more informed deci­sions with greater con­text and speed.

It is the ecosys­tem mod­el we are cre­at­ing — con­nect­ing cus­tomers, work­force, busi­ness­es, and the grid — and our clients are already liv­ing that real­i­ty. Lead­ing util­i­ties such as San Diego Gas & Elec­tric (SDG&E) and many more glob­al ener­gy and util­i­ties providers are a pow­er­ful joint suc­cess exam­ple and sto­ry: through our inte­grat­ed plat­form strate­gic part­ner­ship, they are dri­ving con­tin­u­ous inno­va­tion across dig­i­tal expe­ri­ences and ser­vice mod­ern­iza­tion. Sim­i­lar­ly, at Domin­ion Ener­gy, this vision is reflect­ed in deep focus on redesign­ing the expe­ri­ence around mul­ti-per­sona and tai­lored jour­neys for res­i­den­tial cus­tomers, enter­prise users, land­lords, and agencies.

The val­ue of SEW​.AI and SAP plat­form is that it mul­ti­plies ROI. Our clients don’t have to choose between their enter­prise back­bone and AI-dri­ven ver­ti­cal intel­li­gence. The two deeply ver­ti­cal-focused plat­forms work togeth­er, and that is pre­cise­ly why it is dri­ving this indus­try trans­for­ma­tion at this scale. In a world where util­i­ties are under pres­sure to mod­ern­ize faster than ever, dri­ven by data cen­ter demand, elec­tri­fi­ca­tion, and grid insta­bil­i­ty, hav­ing these two tech stacks work­ing in con­cert is not a lux­u­ry — it is a sur­vival imperative.

Q: You recent­ly intro­duced SEW​.AI COS­MOS Ver­ti­cal AI Native Plat­form. What sparked this vision, and how has it tak­en shape over the past year?

With what we are wit­ness­ing now across the indus­try and in ongo­ing con­ver­sa­tions with ener­gy and util­i­ty lead­ers, part­ners, ana­lysts, and our teams across regions, it became increas­ing­ly clear that the indus­try is at a defin­ing inflec­tion point. Ener­gy and water are no longer back­ground ser­vices; they sit at the cen­ter of eco­nom­ic growth, resilience, and com­mu­ni­ty well-being. At the same time, ris­ing demand, the rapid growth of data cen­ters, the accel­er­a­tion of AI, and mount­ing pres­sure on infra­struc­ture are fun­da­men­tal­ly reshap­ing expec­ta­tions from util­i­ties. It was in this con­text that the vision for SEW​.AI COS­MOS – Ver­ti­cal AI Native Plat­form took shape, not as a response to trends, but as a pur­pose­ful step toward enabling a more con­nect­ed, intel­li­gent, and resilient util­i­ty ecosystem.

When we first spoke about SEW​.AI COS­MOS – Ver­ti­cal AI Native Plat­form at the WE3 Sum­mit, the focus was on what ver­ti­cal AI could mean for this indus­try as an oper­at­ing mod­el. Since then, that vision has steadi­ly matured through real-world plat­form use cas­es across cus­tomer, work­force, and grid trans­for­ma­tion ini­tia­tives. Today, SEW​.AI COS­MOS stands as a uni­fied, ver­ti­cal AI native plat­form that brings togeth­er cus­tomers, work­ers, assets, oper­a­tions, and grid intel­li­gence into a con­tin­u­ous­ly learn­ing ecosys­tem, pur­pose-built for the com­plex­i­ty of ener­gy and utilities.

More than the tech­nol­o­gy itself, it reflects a long-term com­mit­ment to inno­vate with intent and to part­ner close­ly with util­i­ties as they nav­i­gate this transformation.

Con­nect with Deep­ak Garg on LinkedIn. 

Q: SEW​.AI COS­MOS is built around what SEW​.AI calls agen­tic ver­ti­cal AI, which implies the AI is exe­cut­ing, not just rec­om­mend­ing. Where in the util­i­ty work­flow are AI agents cur­rent­ly tak­ing actions with­out human approval, and where have you delib­er­ate­ly kept a human in the loop?

Tra­di­tion­al automa­tion was built around tasks. Agen­tic ver­ti­cal AI is built around out­comes. That dis­tinc­tion mat­ters enor­mous­ly in ener­gy and util­i­ties because util­i­ty oper­a­tions are deeply inter­con­nect­ed. A deci­sion made in cus­tomer oper­a­tions can impact work­force dis­patch. A grid event can impact billing, out­age man­age­ment, and safe­ty oper­a­tions simul­ta­ne­ous­ly. Intel­li­gence can­not remain siloed.

What we are build­ing with SEW​.AI COS­MOS is a plat­form where ver­ti­cal AI can coor­di­nate sig­nals across the enter­prise, iden­ti­fy pat­terns humans can­not process fast enough, and auto­mate low­er-risk oper­a­tional deci­sions in real time. But we are equal­ly delib­er­ate about where human judg­ment remains essential.

In util­i­ties, there are deci­sions that car­ry oper­a­tional, reg­u­la­to­ry, and pub­lic safe­ty con­se­quences. Those should nev­er become black-box automa­tion exer­cis­es. That is why our phi­los­o­phy has always been peo­ple plus AI, not peo­ple ver­sus AI. AI should ele­vate human exper­tise, not remove account­abil­i­ty from the system.

In prac­tice, ver­ti­cal AI can autonomous­ly opti­mize work­flows, pri­or­i­tize oper­a­tional actions, sur­face risk intel­li­gence, and coor­di­nate sys­tem respons­es across inter­con­nect­ed oper­a­tions. But when deci­sions impact safe­ty, crit­i­cal infra­struc­ture, or cus­tomer trust, human over­sight remains essen­tial. The future of util­i­ties will not be ful­ly autonomous sys­tems. It will be intel­li­gent sys­tems where human judg­ment and machine intel­li­gence con­tin­u­ous­ly strength­en one another.

Q: In a world increas­ing­ly dri­ven by hor­i­zon­tal AI plat­forms and gen­er­al-pur­pose intel­li­gence mod­els, why is ver­ti­cal AI becom­ing essen­tial for the ener­gy and util­i­ties industry?

Util­i­ties do not oper­ate in gen­er­al­ized envi­ron­ments. There is a fun­da­men­tal dif­fer­ence between hor­i­zon­tal AI capa­bil­i­ty and ver­ti­cal AI intel­li­gence, and that dif­fer­ence is the entire rea­son SEW​.AI exists. A mod­el that has been trained on util­i­ty-spe­cif­ic data, that does under­stand the physics of pow­er dis­tri­b­u­tion, the reg­u­la­to­ry con­structs of rate cas­es, the oper­a­tional real­i­ty of man­ag­ing aging infra­struc­ture in cli­mate-stressed envi­ron­ments — that mod­el has the depth required to dri­ve real outcomes.

Hor­i­zon­tal AI can­not give you the domain intel­li­gence that comes from years of work­ing inside this indus­try, build­ing mod­els against real util­i­ty data, and under­stand­ing what a ​“five-star” cus­tomer expe­ri­ence looks like oper­a­tional­ly and finan­cial­ly for a reg­u­lat­ed ener­gy or water company.

With SEW​.AI COS­MOS – Ver­ti­cal AI Native Plat­form, we are build­ing the intel­li­gence lay­er that trans­lates raw enter­prise data and AI capa­bil­i­ty into deci­sions and actions that are mean­ing­ful in the con­text of oper­at­ing a grid, man­ag­ing a water­shed, or serv­ing a dat­a­cen­ter. Our role is not to com­pete with gen­er­al­ized plat­forms — it is to unlock the val­ue of both exist­ing and new sys­tems for an indus­try that has some of the most com­plex oper­a­tional require­ments on Earth. A util­i­ty run­ning a grid that must bal­ance sup­ply and demand in real time, respond to storms, meet reg­u­la­tion man­dates, and keep bills afford­able for cus­tomers who are increas­ing­ly finan­cial­ly stressed. That util­i­ty needs a part­ner who under­stands all of that com­plex­i­ty deeply, not a gen­er­al-pur­pose tool that needs to be con­fig­ured from scratch by peo­ple who have nev­er worked in the sector.

Q: You’ve not­ed that 34% of North Amer­i­cans qual­i­fy as low-income, and that bills could rise sig­nif­i­cant­ly as AI-dri­ven demand hits the grid. Can AI itself help keep mod­ern­iza­tion from becom­ing a cost that gets passed through to the cus­tomers least able to absorb it?

This may become one of the defin­ing ques­tions of the ener­gy tran­si­tion. As infra­struc­ture demand accel­er­ates, the tra­di­tion­al response has been straight­for­ward: build more infra­struc­ture and recov­er the cost through ratepay­ers. But the cur­rent scale of mod­ern­iza­tion rais­es a much big­ger ques­tion around afford­abil­i­ty and equi­ty. AI can help change that equa­tion, but only if it is applied intentionally.

We are deliv­er­ing capa­bil­i­ties that proac­tive­ly iden­ti­fy at-risk cus­tomers before they reach a cri­sis point: cus­tomers who are like­ly to dis­con­nect, who qual­i­fy for assis­tance pro­grams they haven’t enrolled in, or who could ben­e­fit from a tar­get­ed effi­cien­cy inter­ven­tion that low­ers their con­sump­tion and their bill. This is ver­ti­cal AI being applied not just to oper­a­tional effi­cien­cy but to social out­comes. For exam­ple, SAP and SEW​.AI are proud to work with util­i­ties like DTE Ener­gy, which is lever­ag­ing tech­nol­o­gy to strength­en engage­ment with vul­ner­a­ble cus­tomers through finan­cial assis­tance and ener­gy afford­abil­i­ty ini­tia­tives. This includes more proac­tive out­reach, improved pledge man­age­ment, and tar­get­ed sup­port to help cus­tomers main­tain con­tin­u­ous util­i­ty service.

Util­i­ties oper­ate with pub­lic trust. The for­ward-look­ing util­i­ties under­stand that mod­ern­iza­tion can­not come at the expense of the com­mu­ni­ties they serve. Intel­li­gence should not only make util­i­ties more effi­cient. It should help make them more equitable.

Q: You’ve described SEW​.AI COS­MOS – Ver­ti­cal AI Native Plat­form as tak­ing the hard path of build­ing for 50 years rather than five quar­ters. What lessons have you learned by tak­ing the hard path?

The first les­son is that depth com­pounds. When you build with gen­uine depth — in domain knowl­edge, in data archi­tec­ture, in reg­u­la­to­ry under­stand­ing, in cus­tomer rela­tion­ships — you cre­ate capa­bil­i­ties that are very dif­fi­cult to repli­cate quick­ly. We made a delib­er­ate choice ear­ly on not to build the prod­uct that was eas­i­est to demo but hard­est to deploy at scale. We built for the oper­a­tional real­i­ty of a util­i­ty: com­plex data envi­ron­ments, lega­cy sys­tems that pre­date the inter­net, work­forces that may be skep­ti­cal of tech­nol­o­gy that doesn’t under­stand their dai­ly real­i­ty, and reg­u­la­tors who need con­fi­dence that AI rec­om­men­da­tions are explain­able and auditable.

That was hard.

The sec­ond les­son is that trust is the only durable cur­ren­cy in this indus­try. Util­i­ties oper­ate crit­i­cal infra­struc­ture that bil­lions of peo­ple depend on for their safe­ty and qual­i­ty of life. They do not give their trust eas­i­ly, and they should not. Every time we have earned that trust, by being hon­est when some­thing didn’t work the way we pro­ject­ed, by stay­ing through imple­men­ta­tion chal­lenges rather than mov­ing on to the next sale, by build­ing rela­tion­ships with oper­a­tors and field crews and not just exec­u­tives, it has opened more doors.

The third les­son, and per­haps the most impor­tant, is that the hard path attracts the right peo­ple. The peo­ple who join SEW​.AI are not here because it was the eas­i­est option. They are here because they believe in what we are build­ing and why. That is what I call the ​“Super War­rior Mind­set” — a mind­set root­ed in own­er­ship, resilience, and the con­vic­tion to solve the hard­est prob­lems fac­ing the indus­try, even when the answers do not exist yet. It is about build­ing with long-term respon­si­bil­i­ty, con­tin­u­ing to inno­vate when the path is uncer­tain, and under­stand­ing that the work mat­ters because mil­lions of peo­ple ulti­mate­ly depend on the sys­tems that we help pow­er and mod­ern­ize. That mind­set has shaped every­thing we have built so far, and I believe it will define every­thing we build in the decades ahead. Build­ing for 50 years means you are always com­mit­ted to some­thing larg­er than the next quarter.

Q: When a util­i­ty CIO is eval­u­at­ing AI part­ners, what are the red flags that a solu­tion is gener­ic tool­ing repack­aged for the indus­try rather than some­thing built with real depth?

The stakes of get­ting this wrong are not abstract. A poor­ly cho­sen AI part­ner can con­sume years of imple­men­ta­tion effort, erode inter­nal con­fi­dence in the tech­nol­o­gy, and leave a util­i­ty fur­ther behind than when it start­ed. So, when I advise CIOs on how to eval­u­ate what they are being shown, I tell them to ask hard ques­tions in four spe­cif­ic areas:

• The pitch leads with tech­nol­o­gy, not out­comes. If an AI provider’s pri­ma­ry ref­er­ence is their mod­el archi­tec­ture, their foun­da­tion­al LLM, or their cloud infra­struc­ture rather than spe­cif­ic util­i­ty prob­lems solved in pro­duc­tion with mea­sur­able results, that is your first sig­nal. Ask direct­ly: where have you been run­ning in pro­duc­tion for more than two years, and what did it cost the util­i­ty before you arrived ver­sus after? If they can­not answer that pre­cise­ly, the depth is not there.
• Ask to speak with oper­a­tions, not just IT. Oper­a­tional staff know with­in min­utes whether a tech­nol­o­gy actu­al­ly under­stands their work. A provider who steers you only toward IT lead­er­ship dur­ing ref­er­ence calls is telling you some­thing. The peo­ple clos­est to the work are the most hon­est eval­u­a­tors of whether an AI solu­tion was built for the industry.
• Their data strat­e­gy is not domain-spe­cif­ic. Ask them to explain what a utility’s data mod­el looks like. How do they han­dle the inte­gra­tion between AMI, CIS, SCA­DA, and work man­age­ment sys­tems? If the answer involves build­ing cus­tom con­nec­tors from scratch for each client, you are look­ing at a hor­i­zon­tal tool try­ing to fit a ver­ti­cal prob­lem. A gen­uine­ly ver­ti­cal plat­form arrives with that inte­gra­tion archi­tec­ture already solved.
• The AI can­not explain its own rea­son­ing. Explain­abil­i­ty in this indus­try is not a nice-to-have — it is a reg­u­la­to­ry and oper­a­tional neces­si­ty. If an AI rec­om­mends defer­ring main­te­nance on a crit­i­cal asset, the oper­a­tor must under­stand why and must be able to chal­lenge that rea­son­ing with their own knowl­edge. Ask the ven­dor direct­ly: can your sys­tem explain, in plain lan­guage, why it made a spe­cif­ic rec­om­men­da­tion? Can that out­put be audit­ed? Can a reg­u­la­tor review the basis for a deci­sion that affect­ed rates or reli­a­bil­i­ty? If the answer is vague, walk away. Black-box AI has no place in infra­struc­ture management.

Q: The work­force con­ver­sa­tion in util­i­ties usu­al­ly cen­ters on giv­ing field crews bet­ter tools. But insti­tu­tion­al knowl­edge is retir­ing out of the indus­try at scale. Can AI real­is­ti­cal­ly cap­ture what a vet­er­an linework­er knows, or is that a dif­fer­ent kind of problem?

This is pre­cise­ly where our peo­ple-plus-AI phi­los­o­phy becomes most tan­gi­ble. Field oper­a­tions are where util­i­ty deci­sions become phys­i­cal, imme­di­ate, and safe­ty-crit­i­cal, and the knowl­edge walk­ing out the door with retir­ing vet­er­ans is not just expe­ri­enced, it is insti­tu­tion­al infra­struc­ture. AI can­not ful­ly replace what a 30-year linework­er car­ries in their instincts. But it can cap­ture pat­terns from decades of work-order his­to­ries, asset behav­ior, and inci­dent data, and put that intel­li­gence direct­ly in the hands of the next gen­er­a­tion of field crews in real time. The goal is not replace­ment. It is amplification. 

In prac­tice, this shows up across the entire field lifecycle:

• Work is pri­or­i­tized not just by urgency, but by com­bin­ing risk, skill fit, and envi­ron­men­tal conditions.
• Crew assign­ment and rout­ing move beyond logis­tics into capa­bil­i­ty match­ing, ensur­ing the com­plex­i­ty of the task aligns with the expe­ri­ence in the field.
• Before a truck even rolls out, pre-job risk intel­li­gence sur­faces poten­tial haz­ards and oper­a­tional constraints.
• Dur­ing exe­cu­tion, field teams are sup­port­ed with real-time, AI-dri­ven guid­ance that flags safe­ty con­di­tions and pro­ce­dur­al steps, while still requir­ing human con­fir­ma­tion for every crit­i­cal action.
• And as con­di­tions evolve, there is con­tin­u­ous coor­di­na­tion between field crews and con­trol cen­ters, ensur­ing deci­sions are made with full sit­u­a­tion­al awareness.

One of the largest util­i­ties in North Amer­i­ca and a long-stand­ing client, Pacif­ic Gas & Elec­tric (PG&E), has a field work­force oper­at­ing across a ser­vice ter­ri­to­ry of extra­or­di­nary scale and com­plex­i­ty. Their work­force is con­nect­ed to our uni­fied work­force expe­ri­ence AI plat­form doc­u­men­ta­tion, oper­a­tional guid­ance, and com­pli­ance con­text direct­ly into the field envi­ron­ment, pro­vid­ing on-the-job train­ing. Not dig­i­tiz­ing knowl­edge for ref­er­ence, but oper­a­tional­iz­ing judg­ment at the point of action, so that insti­tu­tion­al intel­li­gence does not retire when peo­ple do, but con­tin­ues to strength­en how the indus­try is built, main­tained, and secured. And this is what peo­ple plus AI looks like in action.

Q: SEW​.AI oper­ates across dozens of coun­tries. Are there mar­kets where util­i­ties are fur­ther ahead on dig­i­tal matu­ri­ty than their Amer­i­can coun­ter­parts, and what should U.S. util­i­ties be learn­ing from them?

Work­ing along­side 470-plus util­i­ties across 47-plus coun­tries gives us a van­tage point that gen­uine­ly shapes how we build. Every region brings its own objec­tives, its own oppor­tu­ni­ties, and its own pos­si­bil­i­ties. And behind every util­i­ty is a com­mu­ni­ty of peo­ple — bil­lions of peo­ple — who deserve to be engaged, empow­ered, and edu­cat­ed by ener­gy and water that runs their dai­ly lives. That breadth of per­spec­tive shapes how we think, how we build, and what we bring back to every coun­try we oper­ate in. Reg­u­la­to­ry mod­els, cus­tomer expec­ta­tions, and infra­struc­ture real­i­ties in dif­fer­ent coun­tries dic­tate a pace of inno­va­tion that cre­ates lessons worth car­ry­ing back.

For instance, Syd­ney Water, the largest water provider in Aus­tralia, has set a stan­dard for how a util­i­ty can reimag­ine the cus­tomer rela­tion­ship around con­ser­va­tion and dig­i­tal engage­ment, treat­ing water not just as a ser­vice but as a shared responsibility.

In India, Tata Pow­er, one of the most icon­ic brands in the coun­try, embarked on a bold dig­i­tal trans­for­ma­tion jour­ney with SEW​.AI, launch­ing an AI-pow­ered cus­tomer expe­ri­ence plat­form across its mul­ti­ple state dis­tri­b­u­tion com­pa­nies and becom­ing the only util­i­ty in India to launch a Super App that uni­fies ser­vices across all its group com­pa­nies into one seam­less, dig­i­tal-first experience.

In Latin Amer­i­ca, Gas­mig in Brazil is redefin­ing what cus­tomer expe­ri­ence means for the region, deploy­ing our dig­i­tal plat­form includ­ing the ver­ti­cal AI agent built specif­i­cal­ly for ener­gy and util­i­ty use cas­es, from billing and pay­ments to ser­vice requests and dig­i­tal account man­age­ment. These are not pilot pro­grams. These are util­i­ties that made a deci­sion to lead, and they are rais­ing the bar for what the indus­try is capa­ble of deliv­er­ing everywhere.

Vis­it the SEW​.AI web­site. 

Q: Water shows up con­sis­tent­ly in SEW.AI’s mis­sion but gets far less air­time than ener­gy in dig­i­tal trans­for­ma­tion con­ver­sa­tions. Is the water indus­try mean­ing­ful­ly fur­ther behind on dig­i­tal maturity?

Water and ener­gy have always been equal parts of our mis­sion. It is a con­vic­tion we have held since the begin­ning. The con­ver­sa­tion around water in dig­i­tal trans­for­ma­tion cir­cles is qui­eter than ener­gy, but I would not frame it as sim­ply being behind. The water indus­try is nav­i­gat­ing a dif­fer­ent set of pres­sures: aging infra­struc­ture, water scarci­ty, tight­en­ing reg­u­la­to­ry stan­dards, and com­mu­ni­ties that are deeply con­nect­ed to water as a pub­lic good in a way that is dis­tinct from how they relate to ener­gy. What is chang­ing, and chang­ing fast, is the recog­ni­tion that dig­i­tal trans­for­ma­tion in water is no longer option­al. Cli­mate stress, pop­u­la­tion growth, and ris­ing expec­ta­tions from res­i­dents are forc­ing the issue in ways that are accel­er­at­ing the con­ver­sa­tion meaningfully.

Okla­homa City Water Util­i­ties Trust (OCWUT) is a pow­er­ful exam­ple of a water util­i­ty that moved with real clar­i­ty and pur­pose. Uni­fy­ing water, waste­water, and waste into a sin­gle res­i­dent expe­ri­ence, one that is mul­ti­lin­gual, secure, and seam­less, meant dis­man­tling years of frag­ment­ed sys­tems and rebuild­ing around the per­son being served. This work was recent­ly rec­og­nized with an indus­try award, and while awards are not why we build, this one mat­ters because it sig­nals some­thing larg­er: that the water sec­tor is begin­ning to set its own bar for what res­i­dent-first dig­i­tal inno­va­tion looks like. OCWUT did not wait for the ener­gy indus­try to show the way. They defined their own stan­dard, and that is exact­ly the kind of lead­er­ship that moves an entire sec­tor forward.

Q: When you look at the next three to five years — with data-cen­ter demand, elec­tri­fi­ca­tion, and cli­mate volatil­i­ty con­verg­ing on the grid — what would you tell ASUG mem­bers still delib­er­at­ing about where to start?

What makes this moment so extra­or­di­nary for the util­i­ties indus­try is that, for the first time in decades, ener­gy and water have moved to the cen­ter of near­ly every major trans­for­ma­tion hap­pen­ing in the world. The rise of AI, the accel­er­a­tion of elec­tri­fi­ca­tion, the expan­sion of data cen­ters, the growth of dis­trib­uted ener­gy — none of it advances unless the grid evolves along­side it. Util­i­ties are no longer oper­at­ing qui­et­ly in the back­ground of progress; they are becom­ing the infra­struc­ture enabling the next era of human and eco­nom­ic advance­ment. That is what makes this such a defin­ing moment for the indus­try, and hon­est­ly, such a mean­ing­ful time to be build­ing along­side util­i­ty lead­ers who car­ry the respon­si­bil­i­ty of long-term reli­a­bil­i­ty every sin­gle day.

This is also why part­ner­ships like SAP mat­ter so deeply in this trans­for­ma­tion. Util­i­ties already pos­sess decades of oper­a­tional intel­li­gence inside these sys­tems: asset his­to­ries, cus­tomer records, work­force pat­terns, out­age data, oper­a­tional work­flows. That foun­da­tion is incred­i­bly valu­able because it makes trans­for­ma­tion prac­ti­cal, not the­o­ret­i­cal. Util­i­ties do not need to rip apart the sys­tems they trust to run their busi­ness. They need an intel­li­gence lay­er that can con­nect to what already exists, learn from it, and turn it into oper­a­tional deci­sions that teams can act on in real time.

The future will not belong to iso­lat­ed sys­tems or dis­con­nect­ed AI deploy­ments. It will belong to util­i­ties that build con­nect­ed intel­li­gence ecosys­tems, where cus­tomer oper­a­tions, work­force intel­li­gence, grid assets, and busi­ness sys­tems con­tin­u­ous­ly learn from one anoth­er. That is the shift we have been focused on for years: help­ing util­i­ties move from frag­ment­ed oper­a­tions to a mod­el where intel­li­gence flows across the entire val­ue chain. When that hap­pens, AI becomes far more than automa­tion. It becomes the abil­i­ty to antic­i­pate demand, opti­mize ener­gy move­ment, strength­en resilience, and oper­ate the grid with a lev­el of pre­ci­sion and coor­di­na­tion that was pre­vi­ous­ly impossible.