Emerj Artificial Intelligence Research

  • Home
  • Emerj Artificial Intelligence Research

Emerj Artificial Intelligence Research Emerj connects leading AI brands directly with global 2000 enterprise AI buyers - through publishing, media, and exclusive virtual events.

Emerj connects leading AI brands directly with global 2000 enterprise AI buyers - through publishing, media, and a global network of enterprise AI adopters and strategists at companies like Citi, GE, Walmart, Sanofi, and more. Enterprise Leaders: Join over 20,000 Global 2000 leaders and receive our AI ROI research to your inbox every week: emerj.com/n1

AI Vendors: Go to market with confidence and connect directly with enterprise leaders: emerj.com/ad1

Custom roundtables and events: emerj.com/ve1

A chip that's 20% faster on paper can still be the wrong purchase.The number that matters is how it performs on the work...
21/09/2026

A chip that's 20% faster on paper can still be the wrong purchase.

The number that matters is how it performs on the workload you actually run, where the performance difference can be three to five times greater.

In this MIPS-sponsored episode of the AIinBusiness podcast, Drew Barbier, VP of the IP Business Unit at MIPS, joins Emerj's Yolandi de Weerdt to explain why benchmark scores like CoreMark and silicon area can miss the number that should drive a real-time compute decision.

For engineering and product leaders evaluating compute IP for real-time systems: ask vendors for workload-specific performance, not just spec-sheet comparisons, before you commit.

Full episode here:
https://zurl.co/dVHyM

Health systems don't struggle to find AI tools. They struggle to get clinicians to trust and use them at scale. Clevelan...
21/09/2026

Health systems don't struggle to find AI tools. They struggle to get clinicians to trust and use them at scale. Cleveland Clinic's approach shows that adoption is decided by how a tool is selected, sequenced, and judged before it ever goes enterprise-wide.

Emerj's latest analysis examines two clear examples inside Cleveland Clinic's clinical operations:

• Ambient AI documentation — a competitive, multi-vendor pilot gave leadership the evidence to commit to one partner and roll the tool out to thousands of clinicians in months, not years.

• AI-driven sepsis detection — the health system measured success by how much alert noise the model removed, not only by how many more cases it caught, which is what makes clinicians act on an alert.

The principle for other enterprises is that the evaluation design behind an AI deployment shapes its adoption as much as the model itself.

The full article breaks down the business problem, workflow changes, and the maturity of each initiative, including where the published evidence is strong and where it is still self-reported:https://zurl.co/f0uSH

AI trend discovery is not about reading AI headlines.For AI vendors selling into enterprise, the better question is:Wher...
18/09/2026

AI trend discovery is not about reading AI headlines.

For AI vendors selling into enterprise, the better question is:

Where are buyer urgency, capital, and industry pressure starting to converge?

Emerj’s short PDF outlines three practical ways to spot AI trends before a market becomes crowded.

Follow the venture money

Large Series B and C rounds can signal market traction.

The question is not just which AI startups raised capital.

The question is what capabilities they are enabling, what data they require, and what success claims they can prove.

Listen to leading industry CEOs

Press releases can be poor signals.

Executive panels, speeches, and public comments from CEOs and COOs can reveal what industry leaders see as future risks, competitive threats, and strategic priorities.

Lock on to industry priorities

Enterprise AI adoption rarely starts with novelty.

It gains traction where AI aligns with existing business pressure.

In pharma, that might mean drug development.

In banking, post-2008 regulatory pressure helped make fraud and compliance stronger areas for AI investment.

The PDF is a quick guide for using secondary research to identify AI opportunities in any sector.

Access the full guide here:
https://zurl.co/7UeHf

18/09/2026

The flashiest AI use case is rarely the best place to start.

Xiong Liu funds the foundation first, then scales everything else off of it.

In this CDD Vault-sponsored conversation, Xiong Liu, Director of Data Science & AI at Novartis, joins Emerj’s Yolandi de Weerdt to lay out the sequencing rule behind sustained AI adoption:

Build the data foundation once. Then earn the green light to scale through proven adoption.

For R&D leaders sequencing next year’s AI roadmap:

Before funding the next use case, confirm the data foundation and semantic layer it depends on already exist.

The roadmap decision may depend less on which use case comes first and more on whether the foundation underneath it is ready.

Worth your 45 minutes: https://zurl.co/q4ggL

Healthcare organizations improve outcomes when they can deliver confident, consistent decisions. AI earns a place by rem...
17/09/2026

Healthcare organizations improve outcomes when they can deliver confident, consistent decisions.

AI earns a place by removing diagnostic ambiguity — turning fragmented cardiac data into clarity clinicians can act on.

In this Cleerly‑sponsored episode of the AI in Business podcast, Jim Hartman, Chief Commercial Officer at Cleerly, joins Emerj’s Marilie Fouché to examine how AI can help close that gap.

The conversation highlights a central challenge for hospitals:

Ambiguity between tests and the clinical decision delays accurate diagnosis and complicates care pathways.

Jim outlines three areas health systems evaluate when considering AI for cardiac imaging:

• Clinical impact

• Workflow efficiency

• Financial return

Jim offers leaders a clear way to evaluate AI in the full episode:
https://zurl.co/yUorl

17/09/2026

Twenty percent faster on a spec sheet.

Three to five times faster on the workload that actually matters.

In this MIPS-sponsored episode of the AIinBusiness podcast, Drew Barbier, VP of the IP Business Unit at MIPS, tells Emerj's Yolandi de Weerdt why generic benchmarks like CoreMark and raw IPC scores can undersell what workload-specific compute IP can deliver for real-time AI systems in vehicles, robots, and industrial equipment.

For leaders sourcing compute for these systems, the spec-sheet number and the number that matters aren't the same.

Ask vendors to show you performance on the workload you need to run before you commit.

Full episode here:
https://zurl.co/aEenN

A model can reach state-of-the-art accuracy and still deliver zero business value.Xiong Liu says the missing link is a n...
16/09/2026

A model can reach state-of-the-art accuracy and still deliver zero business value.

Xiong Liu says the missing link is a named stakeholder decision.

In this CDD Vault-sponsored conversation, Xiong Liu, Director of Data Science & AI at Novartis, tells Emerj’s Yolandi de Weerdt why downstream value is difficult to measure when no stakeholder decision is attached to the model.

For AI and data science leaders under pressure to prove ROI to budget owners:

Before scaling a model, confirm which stakeholder decision its output feeds.

If that decision is unclear, the business case for scaling is unclear too.

Explore the full breakdown: https://zurl.co/MneB3

16/09/2026

Companies deploy AI agents and call the project done.

Vanessa Tabbert took a different approach.

In this Salesforce-sponsored episode of the AI in Business podcast, Vanessa Tabbert, VP of Agentic Transformation and Sales Development at Salesforce, joins Emerj’s Yolandi de Weerdt to explain how she put her AI SDR agent on her org chart and coached it like a rep, not a tool.

For sales and revenue operations leaders running lean teams, that changes the management question.

The value of an AI agent doesn’t end at deployment.

It requires feedback, coaching, and ongoing management.

After one month of structured feedback, Vanessa’s team went from booking 150 meetings a month to 150 a week.

That points to a different way to think about AI agents:

Not as software you deploy.

As capacity you manage.

Full episode: https://zurl.co/PH9tE

LLM pilots are proving useful. The harder question is what happens when 200 users become 20,000.When two use cases becom...
16/09/2026

LLM pilots are proving useful. The harder question is what happens when 200 users become 20,000.

When two use cases become 2,000, infrastructure becomes a business continuity decision.

And if the business depends on a vendor whose commercial terms could change, that dependency becomes a strategic risk.

Rodrigo Liang, Co-founder and CEO of SambaNova, joins Emerj's Daniel Faggella on the AI Infrastructure Podcast to unpack why enterprises getting this right are building for sovereignty from the start.

That means breadth across models, clouds, and chip architectures, so a service never depends on a single vendor to keep running.

Liang also explains:

• Why regulated firms are keeping customer data inside their own firewalls until regulation settles

• Why the power a legacy data centre can draw now determines which infrastructure it can run

• Why he tells CIOs to start with code generation, the one at-scale use case nearly every company has, and a fast way to learn what AI at scale will cost

• Why business continuity, rather than benchmark performance, is driving multi-vendor infrastructure strategies

• How agentic systems broken into checkable steps give regulated firms the traceability they need

• Why assuming 100% of the business will be AI-enabled changes the infrastructure choices made today

The full episode gives enterprise leaders a clearer view of what AI infrastructure needs to support as adoption moves from pilots to business-critical scale.

https://zurl.co/bVGl9

If your primary AI vendor changed its terms tomorrow, could your business keep running?

A commercial plan signed off at head office and what happens on a store shelf are two different things — and for consume...
15/09/2026

A commercial plan signed off at head office and what happens on a store shelf are two different things — and for consumer goods companies, the distance between them is one of the most expensive disconnects in the business.

Stephanie Lilley, VP of Transformation for Europe at Reckitt, joins the AI in Business Podcast to unpack why that gap persists: too many hands touch the plan, ex*****on is hard to measure, and traditional rep call cycles set priorities four to six weeks before anyone walks into a store.

Lilley joins host Daniel Faggella, Emerj CEO and Head of Research, to walk through where AI is beginning to change the model — from image recognition that verifies compliance and opens the door to crowdsourced merchandising, to dynamic routing that sends reps to the stores where the return on their day is highest, to LLMs that surface the insight field teams collect every day but rarely get to use.

The full episode gives leaders a clear picture of:

Why the plan-to-shelf gap costs CPGs on both the investment and the revenue side
How to balance dynamic, ROI-driven routing with the relationship-building that stores still need
Why the single biggest determinant of success isn't geography or field model, but how good and how recent your store-level data is

https://zurl.co/DkZst

Is your field team executing yesterday's plan — or reacting to what your data is telling you today?

Address

MA

Telephone

+16179458567

Website

, ,

Alerts

Be the first to know and let us send you an email when Emerj Artificial Intelligence Research posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Contact The Business

Send a message to Emerj Artificial Intelligence Research:

Shortcuts

  • Want your business to be the top-listed advertising & marketing company?

Share