Sugumaran Balasubramaniyan

AI/ML Engineer | Professor @SKEMA_BS | LLM Systems, Agentic AI & MLOps | AWS, Docker, Kubernetes | Building production-grade AI systems

[Sugumaran Balasubramaniyan @sugumaran___]

We are at a massive inflection point in AI deployment.

Building autonomous agents directly from the terminal gives us the raw capabilities, but bridging the gap from local development to enterprise-scale execution is where the real industry shifts happen.

That is why I am heading to Paris on 8-9 July 2026.

I am incredibly proud to announce that I am an Official Ambassador for RAISE Summit 2026.

This is not just another tech meetup; it is the world's leading AI event where visionaries and builders actually converge.

The scale of this ecosystem is unprecedented:
• 9,000+ global leaders in the room
• 80% C-level decision-makers
• 40% representation from Fortune 1000 companies
• 350+ speakers & 200+ top-tier exhibitors

Most importantly, it is backed by the massive infrastructure and compute giants scaling our models: @Google, @awscloud , @nvidia , @MistralAI , and @GroqInc

If you are serious about building the future of AI whether that is open-source agentic systems or massive enterprise LLM deployments you need to be in the room. Let's connect at the Carrousel du Louvre.

🎟️ Use my personal ambassador code SUGAMB26 to claim a 20% discount on your pass.

🔗 Secure your ticket here: https://t.co/biVM5DWbFL 

@RAISESummit

#RAISE2026 #RAISESummit #RAISEWeek

[RAISE Summit @RaiseSummit]

Stevie Case (@KillCreek )(Vanta @TrustVanta ) on the RAISE stage, on a shift happening faster than the job titles can keep up:

Stevie Case: "Just in the last three months, we have seen a 300% increase in functional engineering roles."

Go-to-market engineers, legal engineers, marketing engineers, whole categories of work that didn't exist a year ago. It isn't just upskilling, she argued, it's a wave of new roles being invented in real time, and companies hiring people into jobs the role itself has never had before.

Proof that AI isn't only changing how we work, but what the work even is. RAISE Summit 2026, Paris.

#RAISE2026 #AI #Paris

[RAISE Summit @RaiseSummit]

The results are in. RAISE the Stakes 2026, Europe's largest AI startup competition, has its winners, crowned live on the Master Stage after the final pitch.

🥇 First place: QurieGen, led by CEO and founder @KingaMatula .
🥈 Second place: Strawberry Browser, @charles_maddock , CEO.
🥉 Third place: @NeuralTrustAI , Rodrigo F. Baon, CMO.

Three standout teams from a field that started at ten. Congratulations to all of them, and to Kinga and QurieGen for taking the title.

This is what raising the stakes looks like.
RAISE Summit 2026, Paris.

#RAISE2026 #AI #RAISEtheStakes #Paris

[RAISE Summit @RaiseSummit]

Late on the Master Stage, two firesides split the same question in two: what does it take to build AI at scale, and what does it take to make it pay?

First, the infrastructure. In Scaling AI Infrastructure for the Agentic Era, @CharlieKawwas (@Broadcom ) sat with Tony Kim (@BlackRock ) to explain why the biggest labs are walking away from general-purpose compute. His metaphor did the work.

Charlie Kawwas: "This is why people come to the haute couture. You come to us. You tell us what kind of tuxedo or dress you want. We will customize it and build it for you and remove that general compute tax."

That tax, he argued, is paid twice, in efficiency and in margin, which is why four of the five hyperscalers now spending most of the world's compute budget are moving custom at the rack level. The demand behind it is staggering: from under $1 billion in this space two years ago to a projected $100 billion-plus, and labs scaling from one or two gigawatts toward more than ten. Whoever first integrates GPU, XPU and CPU into a single platform, he suggested, may reach AGI first.

Then, the payoff. In his fireside, Mark Hura (@Oracle ), with Alex Banks (@thealexbanks ) (The Signal @thesignalAI_ ), moved the conversation from stacks to outcomes.

Mark Hura: "They're not thinking about buying an AI stack. They're looking at how do we bring an outcome."

Customers have crossed from experimenting to executing, he said, and the wins are concrete: AI cutting construction billing errors by over 72% and removing 130,000 support calls a year, clinicians handed back time through assistants that prep every patient summary. The real unlock, in his view, is private enterprise data, where the IP lives, and the rule is simple: start with the business problem, and the technology follows.

Custom silicon below. Business outcomes above. The full stack of turning intelligence into results, in twenty-minute windows.

#RAISE2026 #AI #Paris

[Sugumaran Balasubramaniyan @sugumaran___]

"Powering AI Infrastructure: Solving the 21st Century Manhattan Project."

Scaling agentic systems isn't just a software problem, it's a massive physical constraint. The sheer density of networking and energy required for production-grade AI is unprecedented.

Insightful panel at the RAISE Summit right now with:
Ben Richardson of @CoreWeave
Shaun O'Meara @ShaunDuvin of @MirantisIT
Alex Saroyan @alex_saroyan of @netris_io
June Paik @junehopaik of @FuriosaAI
Vik Malyala @malyala of @Supermicro
Jeremie Eliahou Ontiveros @JeremieEO of @SemiAnalysis_

@RaiseSummit @amarmic @HenriDelahaye

#RAISESummit #AIInfrastructure #MLOps #EnterpriseAI

[Sugumaran Balasubramaniyan @sugumaran___]

The capability overhang in AI is real, and bridging it requires serious infrastructure.

We cannot scale agentic workflows into production without fundamentally rethinking the hardware layer. Tuning in to "Closing the Capability Overhang: Compute for the Agentic Era" at the RAISE Summit to hear how leaders are tackling industrial-scale compute constraints.

Insights from:
Sachin Katti (@sk7037) of @OpenAI and @andrewdfeldman of @cerebras  
Henri Delahaye of @RaiseSummit

#RAISESummit #AIInfrastructure #MLOps #EnterpriseAI #AgenticAI

[Sugumaran Balasubramaniyan @sugumaran___]

"Agents Are Here. Compute Has to Change." ⚡️

Scaling autonomous systems is pushing our infrastructure to the absolute limit. We can't build reliable, production-grade agentic workflows without fundamentally rethinking the hardware layer to support them.

Fascinating insights from Sachin Katti @sk7037 , Head of Industrial Compute at @OpenAI, on the realities of industrial-scale AI today at the RAISE Summit. 🇫🇷

@RaiseSummit @amarmic @HenriDelahaye

#RAISESummit #AIInfrastructure #MLOps #EnterpriseAI #OpenAI

[Sugumaran Balasubramaniyan @sugumaran___]

Democratizing engineering through AI is incredible, but it introduces a massive governance challenge.

If agentic tools are empowering more teams to build, we need even stricter evaluation harnesses to ensure production reliability.

Catching "The Democratization of Engineering" at the RAISE Summit with Sassine Ghazi @SassineGhazi of @Synopsys and @ArjunKharpal of @CNBC.

@RaiseSummit @amarmic @HenriDelahaye

#RAISESummit #MLOps #AgenticAI #ProductionML

[Sugumaran Balasubramaniyan @sugumaran___]

The physical footprint of our models is the next major bottleneck in production AI.

Scaling agentic systems means scaling compute, which demands a serious conversation about power and sustainability.

Happening right now at @RaiseSummit: diving into "Crusoe's Energy-First Vision for the AI Cloud" with Chase Lochmiller @ChaseLochmiller of Crusoe and Henri Delahaye of @RaiseSummit

Presented by @cognition_labs. 🇫🇷

@amarmic

[Sugumaran Balasubramaniyan @sugumaran___]

The hardware layer is the unsung hero of production AI. ⚡️

You can't build reliable, multi-step agentic systems without the compute infrastructure to back them up. Right now at the RAISE Summit, we are diving deep into the silicon and strategy side with the 5C x AMD fireside chat.

Critical infrastructure insights from:
David Bitton of @5CGroupAI
Andrew Dieckmann of @AMD
David Dubrovsky of @5CGroupAI
Moderated by Rachel Metz of @business

Presented by @cognition_labs. 🇫🇷

@RaiseSummit @amarmic

#RAISESummit #MLOps #AIInfrastructure #EnterpriseAI

[RAISE Summit @RaiseSummit]

Day 2 opens with @Google .

Amin Vahdat, SVP and Chief Technologist of AI and Infrastructure, took the Master Stage to open the final day with a keynote on the foundations: The Physics of Thought and the Architecture of Intelligence.

His frame for the whole era:
Amin Vahdat: "The industrial revolution was the muscle multiplier. We're now in the era of intelligence, the mind multiplier."

The biggest day, underway. RAISE Summit 2026, Paris.

#RAISE2026 #AI #Google #Paris

[Sugumaran Balasubramaniyan @sugumaran___]

The real test of an AI system isn't how it demos, but how it executes in production.

Fantastic insights from the "Building AI that gets work done" session at the RAISE Summit with Gautier Cloix (@GautierCloix) of @hcompany_ai and Nico Rosberg (@NicoRosberg) Formula 1 @F1 world champion, @RosbergVentures, presented by @cognition_labs.

@RaiseSummit @amarmic

#RAISESummit #EnterpriseAI #MLOps #AgenticAI

[RAISE Summit @RaiseSummit]

Day 2's morning ran on one throughline: the models aren't the moat. What you build on top of them is.

Grant Lee (@thisisgrantlee ) (Gamma @GammaApp ), with Mike Butcher ✍️ (@Pathfounders ), made the sharpest version of the case.

Grant Lee (@thisisgrantlee ): "AI is the steel of our generation. The question is who turns this general-purpose technology into purpose-built technology."

His warning: most AI today is a bolt-on, steel used to make a stronger horse carriage, when the real opportunity is to invent the car. Models are interchangeable and switching costs are low, so the edge goes to whoever builds something purpose-built on top. He pushed back on the everyone-must-be-a-builder orthodoxy too, arguing AI should free salespeople and support teams for the human relationship, not replace it.

Ariel Cohen (@arielcoco ) (@Navan ), with @MollySOShea (@sourceryy ), took that human thread further, straight into a room full of AI companies.

Ariel Cohen (@arielcoco ): "I'm bullish on humans, more than ever. And I'm perfectly happy saying that at an AI conference, as an AI company."

Travel is too complex and too personal to fully automate, he argued, which is why Navan built an orchestration layer between AI agents and real travel agents. The proof: their chatbot Eva already handles 60% of support at near-human satisfaction, while the system saves companies around 15% of their travel budget. You cannot hallucinate an upgrade, as he put it, so the human stays in the loop where it counts.

Then Olivier Pomel (@oliveur ) (Datadog @datadoghq ), with Karim Jalbout (Konstellation Advisory), grounded it in reliability. Enterprises won't go all-in on one model, he noted, they run a mix and need it observed, secured and verified, because self-healing software is harder than self-driving, and a machine that's wrong twice loses the customer for good.

The models commoditize. The moat is what you build, how well it runs, and knowing where humans still belong.

#RAISE2026 #AI #Paris

[Sugumaran Balasubramaniyan @sugumaran___]

Stop hyping raw model capabilities. Start measuring business value.

We just had the most critical conversation in enterprise AI at the RAISE Summit: "Cracking the ROI Code: Is Intelligence Meeting the Income Statement?"

Enterprise AI doesn't need flashier models; it needs a strict harness, systematic evaluation, and governance to actually deliver reliable ROI.

Incredible insights on the realities of production ML from this powerhouse panel: 
May Habib (@may_habib) of Writer
Vishal Talwar of @FedEx
Amit Zavery (@azavery) of @ServiceNow
Stevie Case (@KillCreek) of Vanta
Rick McConnell of @dynatrace
Moderated by Karen Kwok (@karenkkwok) of @Reuters

Building for production > Building for demos. 🇫🇷🛠️

@RaiseSummit @amarmic

[RAISE Summit @RaiseSummit]

RAISE Summit 2026, Day 1 is done. The room at the Carrousel du Louvre went quiet hours ago, Paris didn't.

Dinners running late. Founders and investors still deep in it across the city tonight. RAISE doesn't end when the doors close. It just changes address.

Tomorrow, round two.

#RAISESummit #RAISE

[RAISE Summit @RaiseSummit]

Between the stages, the summit keeps moving.

Walk the RAISE floor and this is what you find: booths in full swing, live product moments, and the between-sessions conversations where introductions turn into partnerships.

The talks set the agenda. The floor is where people actually connect.

RAISE Summit 2026, Paris. Day 1.
#RAISE2026 #AI #Paris

[RAISE Summit @RaiseSummit]

Announced live from RAISE: @SambaNovaAI has raised $1 billion at an $11 billion valuation.

CEO and co-founder Rodrigo Liang (@RodrigoLiang) broke the news on stage. The Series F, a first close led by @generalatlantic with @IntelCapital, @BlackRock and the Qatar Investment Authority joining, roughly quintuples the company's valuation in a matter of months, a bet on SambaNova as a serious challenger to NVIDIA in AI inference.

And it lands with a marquee proof point: JPMorgan Chase has selected SambaNova as its inference-infrastructure partner, powering secure, on-premises AI inside the bank.

Liang's read on the moment: inference has broken everything open, and a standalone company can now move fast enough to seize it.

Big news, live from the RAISE stage. Day 1, Paris.

#RAISE2026 #AI #SambaNova #Paris

[Sugumaran Balasubramaniyan @sugumaran___]

Taking a break from orchestrating agentic systems to enjoy the evening here in Paris.

Representing the @RaiseSummit as an Official Ambassador this week has been incredible, and tonight's "Raise your Glass with VivaTech" cocktail dinner is the perfect way to cap off the day.

Always a pleasure connecting with the European AI ecosystem. 🥂🇫🇷

@VivaTech

[SKEMA Business School @SKEMA_BS]

[ACTUALITÉ 🚀] La taxe d’apprentissage joue un rôle essentiel dans le développement et la croissance de SKEMA et lui permet de continuer à évoluer. Consultez la plateforme SOLTéA pour affecter votre solde à SKEMA: https://t.co/qE9g0HPDWz #WeAreSKEMA

[RAISE Summit @RaiseSummit]

The earlier sessions argued that intelligence has to get cheaper. These two showed the two places that actually happens: the factory floor underneath it, and the creative floor on top.

First, the machine. Giordano Albertazzi (@Vertiv ) reframed the AI data center as something you no longer build, but manufacture.

Giordano Albertazzi: "We have to move from building a data center to manufacturing one, designing and optimizing the entire data center as a single product."

His measure of success is the one echoing across the summit, output per megawatt, tokens per megawatt, and his warning was that a single unoptimized seam is a capacity leak that breaks the token economics. With rack density climbing from tens of kilowatts toward a megawatt per rack, he made the case for fully modular, converged infrastructure: half the deployment time, a smaller footprint, and a lower cost per token as the payoff.

Then, the human. On the Creativity in the Age of AI stage, @theduncanclark (@canva ) flipped the automation question, asking not who you can replace but what your team could do if AI properly amplified them, a shift that now has a quarter of a billion people designing every month. Guillaume Princen (@G_Princen ) (@AnthropicAI ) named the principle underneath it.

Guillaume Princen: "AI should help you think further, not replace your thinking."

It's why Anthropic keeps ads out of Claude, he explained: a tool built to sharpen thinking, not to farm attention. And Europe is leaning in, having gone from four countries in the world's top 20 for AI usage to ten in six months, now half the list.

Megawatts below, imagination above. The same system, seen from both ends.

#RAISE2026 #AI #Paris

[Sugumaran Balasubramaniyan @sugumaran___]

"Inference's Scaling Era" live at #RAISESummit2026.

The training hype is fading; it's all about inference economics now. If you want to scale models in production, optimizing KV cache and automating hardware provisioning is non-negotiable.

Essential insights from Tuhin Srivastava (CEO of @baseten ) and Corinne Riley (@GreylockVC ).

@RaiseSummit @amarmic