In this edition of Nu-Pie Pulse:

Most companies investing heavily in marketing are flying partially blind numbers scattered across five platforms, assembled manually every Monday. There is a better way, and it is already running.

Nu-Pie built a fully automated marketing metrics pipeline for a client investing across LinkedIn, email newsletter, and YouTube. The goal: zero manual data collection, one click to refresh, every Monday morning.

A scheduled Power Automate flow triggers every Monday at 7 AM. It calls each platform's API, extracts the week's performance data, appends it to a centralised Excel tracker on SharePoint, and triggers a Power BI dataset refresh. By 8 AM, the dashboard is current without a single person touching a file.

The architecture is built on tools the client already owned: Power Automate, SharePoint, Excel Online, and Power BI Service. No new software licences. No external vendors. The automation runs inside the organisation's existing Microsoft 365 environment.

The near-term question is not whether marketing automation will become standard practice it already is in organisations that have the infrastructure in place. The question is whether your marketing and data teams are structured to build it, and whether your leadership is seeing the output it enables.

Teams that replace manual reporting with automated pipelines do not just save time. They change the quality of decisions being made because insights arrive earlier, more consistently, and without the errors that accumulate when humans touch spreadsheets every week.

If your marketing dashboard is being assembled by hand, that is a solvable problem. It was harder to solve in 2023. It is not in 2026.

Big Pharma's AI Arms Race, This Week

What's Happened:

Merck announced a landmark multi-year partnership with Google Cloud worth up to $1 billion, deploying an agentic AI platform powered by Gemini Enterprise across its entire enterprise. Coverage spans R&D, manufacturing, commercial and corporate functions, with Google Cloud engineers embedded directly alongside Merck teams. Full rollout is planned by end of 2026.

Why It Matters:

This goes far beyond typical pharma/AI deals focused on discovery alone. Merck is betting on AI as operational infrastructure not just a research tool. The timing is notable: it is simultaneously laying off thousands of staff to save $3B annually by 2027. AI isn't arriving alongside the workforce in many functions, it may be replacing it outright.

"By deploying an industry-first agentic ecosystem powered by Gemini Enterprise, Merck is building a future where the speed of AI and the expertise of human ingenuity come together."
Thomas Kurian, CEO, Google Cloud

Nu-Pie's Take:

This is a clear signal that data infrastructure is becoming the actual drug pipeline. When a company like Merck signs a billion-dollar dependency and embeds a vendor's engineers into its walls, it's not buying software it's restructuring its operating model around AI outputs. The analytics layer is now load-bearing. Teams that can't translate model outputs into regulatory-grade decisions will be the bottleneck, and that's precisely the gap where specialized AI analytics firms are built to operate. Watch for Pfizer and AstraZeneca to make similar moves before Q3.

Source: Merck.com | Fierce Pharma | BioPharma Dive

What's Happened:

Eli Lilly signed a multi-program strategic partnership with Profluent, an AI startup specializing in protein-design foundation models, worth up to $2.25 billion in milestones and royalties. Profluent's AI designs custom site-specific recombinases enzymes that can insert entire kilobase-scale stretches of DNA at precise genome locations targeting diseases where CRISPR and base editing fall short.

Why It Matters:

Kilobase-scale gene insertion is considered genetic medicine's "holy grail." If AI-designed recombinases can reliably deliver large genes, diseases driven by hundreds of different mutations where no single edit works across patients could be treated with one therapy. That's a step-change in both clinical economics and addressable patient reach.

"Many genetic diseases involve hundreds of different mutations across patients no single edit can address them all. Inserting a complete functional gene opens the potential to address diseases with high mutational heterogeneity using a single therapeutic." Profluent

Nu-Pie's Take:

AI's most disruptive role in pharma isn't search or summarization it's generative biology. Profluent isn't analyzing existing molecules; it's designing entirely new ones that wouldn't exist without the model. From an analytics standpoint, the bottleneck is shifting from data collection to experimental validation. The companies that win here will be those that close the loop between in silico model outputs and in vivo wet-lab feedback making AI not just a discovery tool, but a continuous learning system inside the pipeline.

Source: Business Wire | STAT News | pharmaphorum

What's Happened:

Anthropic acquired Coefficient Bio a stealth AI drug-discovery startup with fewer than 10 employees for approximately $400 million in an all-stock deal (April 3, 2026). The ex-Genentech co-founders join Anthropic's healthcare group, bringing AI tools for drug R&D planning, clinical regulatory strategy, and candidate identification, integrated with Benchling, PubMed, and 10x Genomics.

Why It Matters:

This is Anthropic's first major acquisition and makes it the only major AI lab with in-house pharmaceutical domain expertise embedded directly in its model team. For pharma procurement teams deciding which AI partner to trust with molecule-level R&D, Genentech pedigree is a trust signal no benchmark sheet can replicate. Analysts are already calling it the first domino expect competing frontier labs to respond with similar biotech acqui-hires.

"It's the first acquisition of a tech-bio company in a long time. It demonstrates how the thinking is changing in terms of allocating capital around AI and drug development and definitely sets a precedent."
Orr Inbar, CEO, QuantHealth

Nu-Pie's Take:

The $400M price tag is misleading framing Anthropic isn't buying a product, it's buying domain credibility at the moment pharma is asking its hardest question: which model actually understands our data? General-purpose LLMs hit a wall when the context is a Phase II protocol or a regulatory dossier. Coefficient Bio's founders close that gap. The takeaway for analytics companies in life sciences is clear: vertical depth now outweighs horizontal reach. The next moat in pharma AI isn't the biggest model it's the one that already knows what an IND looks like.

Source: TechCrunch | PharmaVoice | Fierce Biotech

Subscribe

Each edition, we deliver the best of analytics, AI, and human brilliance in a crisp, curated edition.

Let’s decode data. Let’s build what’s next. Let’s do it - together.

- Team Nu-Pie

Think Analytics. Think Nu-Pie.

Be Part of Datacon

Let’s stay in touch!