Graph of Nodes and Edges

Looking to get moving with AI? Here are six New Year’s resolutions for biotech leaders—moving from personal experiments to strategic advantage, starting with things you can try this week.

If you’ve been watching the AI landscape evolve and wondering when the right time to act is—good news: it’s now, and you haven’t missed anything.

The past two years have given us something valuable: hindsight. We’ve seen what works, what doesn’t, and where the real opportunities lie. The hype cycle has settled enough that practical applications are becoming clear. The tools have matured. And for biotech leaders ready to move, the path forward is more accessible than ever.

Here’s our take on six ways for biotech leaders to put the cart in motion as we head into 2026—starting with things you can try yourself, building toward capabilities for your team, and setting the stage for a more deliberate approach down the road.

Here’s what we’ll cover:

  1. Get Claude and Teach It a New Skill — Build a personalized AI agent without writing a single line of code
  2. Automate a Task You Never Make Time For — Use tools you probably already have to eliminate operational drag
  3. Let AI Wrangle Your Company Knowledge — Structure your data so your team can find it and AI can actually use it
  4. Activate Data to Augment Your Decisions — Tap into the public data you may be ignoring
  5. Understand Your Company’s “Quiet Pilots” — Learn from the AI experiments already happening
  6. Take Time to Define Your Roadmap — Move from AI as a useful tool to a strategic asset

1. Get Claude and Teach It a New Skill

If you haven’t explored Claude’s desktop application yet, make 2026 the year you do. It’s arguably the easiest way to build a powerful, personalized AI agent—without writing a single line of Python.

The secret? Skills.

A skill is essentially a set of instructions that teaches Claude how to excel at a specific type of task. Think of it as giving Claude a playbook: when you ask it to do something covered by that skill, it follows your carefully defined best practices rather than improvising from scratch. Skills live as simple folders on your computer, and Claude reads them automatically when relevant.

Start personal if you want—teach Claude how to organize your chaotic Downloads folder, triage your desktop into a sensible file structure, or format notes the way you actually like them. Once you see how quickly you can encode your preferences into something reusable, you’ll start seeing opportunities everywhere.

Then bring it to work. Here are just a few examples where skills get powerful for biotech teams:

  • Literature triage skill — Teach Claude how you want scientific papers evaluated: what to extract, how to score relevance, which journals matter most for your therapeutic area
  • Protocol review skill — Encode your internal checklist for reviewing experimental protocols, flagging common issues, or comparing against internal templates
  • Vendor evaluation skill — Give Claude your criteria for assessing CROs, CMOs, or technology partners so it can help you structure diligence consistently

The barrier to entry is remarkably low. You’re not building software—you’re writing down expertise in a structured way. And if you’re not sure where to start, Claude has a built-in skill-creator skill that walks you through developing your own. Anthropic also publishes detailed documentation and examples to help you get going.

Pair skills with connectors to your file system, databases, or internal tools, and you’ve built a genuinely useful agent tailored to how your team works—no engineering backlog required.

2. Automate a Task You Never Make Time For

Every organization has them: tasks that matter, but never feel urgent enough to fix. They sit on the backlog, quarter after quarter, while your team burns hours on manual workarounds.

Make 2026 the year you finally knock one off the list.

The good news? You don’t need a major IT initiative to make progress. If you’re already in Microsoft 365, you have (or can get) more automation power than you’re probably using. Power Automate and Copilot Studio can handle the kinds of operational tasks that quietly drain time across your organization:

  • Routing contracts through approval workflows and collecting signatures automatically
  • Triaging incoming invoices, matching them to POs, and flagging exceptions for review
  • Sending reminders when documents are expiring—vendor agreements, training certifications, equipment maintenance schedules
  • Auto-filing email attachments into the right SharePoint folders based on sender or content

These aren’t glamorous AI use cases, but they’re high-value. They reduce errors, eliminate bottlenecks, and free up your ops team to focus on work that actually requires judgment.

3. Let AI Wrangle Your Company Knowledge

Biotech companies do extraordinary science. But behind the scenes, (we know from experience) that data and knowledge are often scattered across SharePoint sites, buried in email threads, or trapped in slide decks from five years ago.

This fragmentation creates problems today: teams waste time hunting for information, decisions get made without the full picture, and institutional knowledge walks out the door when people leave.

But here’s the bigger risk: fragmented knowledge limits what AI can do for you. The most powerful applications of generative AI depend on context—the ability to ground responses in your data, your documents, your decisions. If that context is scattered and inaccessible, you’re capping your upside before you even start.

Make 2026 the year you clean things up.

This doesn’t require a major initiative—or a specialized tool. Start where you are:

  • Point Copilot at the right sources. If you’re using Microsoft 365, simply connecting Copilot to your most important SharePoint sites can unlock meaningful value. It’s not magic, but it’s a real step toward making institutional knowledge accessible.

  • Build a basic structure for context engineering. Set up a simple database—graph or relational—that captures the entities and relationships that matter to your business: programs, targets, partners, decisions, documents. This becomes the foundation for grounding AI in your specific context.

  • Connect it to your team. Use a simple connector—like a free, open-source MCP server—to give your team direct access to that structured knowledge through the AI tools they’re already using.

You don’t need to buy anything new for this. The databases you need are free. The data belongs to you. And chances are, you’ve already bought the brain—Copilot, Claude, or ChatGPT—that can bring it all together.

4. Activate Data to Augment Your Decisions

There’s a staggering amount of free, public data relevant to your pipeline—and most biotech companies barely touch it.

ClinicalTrials.gov. FDA adverse event databases. Patent filings. PubMed. OpenTargets. ChEMBL. Gene expression repositories. Exposed APIs from organizations that actually want you to use their data.

This isn’t hidden information. It’s just underutilized. And too often, the assumption is that integrating it requires a massive data science team or a platform investment you can’t afford.

It doesn’t.

We’ve already talked about what you can do with Claude and no code. With just a little bit of code, the possibilities expand dramatically. You don’t need to be Recursion to integrate public data into your decisions. A few API calls and a well-structured prompt can help you:

  • Scan competitive clinical trial activity and surface signals relevant to your program
  • Pull adverse event patterns from FAERS to inform your safety strategy
  • Cross-reference target biology across multiple databases to pressure-test your thesis
  • Monitor patent landscapes for freedom-to-operate insights

The data is there. The tools to query it are accessible. The models to synthesize it are already on your desktop. What’s missing is usually just the intention to connect the dots.

Make 2026 the year you start treating public data as a strategic asset—not something you check occasionally, but something that actively informs how you prioritize, plan, and execute.

5. Understand Your Company’s “Quiet Pilots”

Here’s a reality check: even if your company doesn’t have a formal AI strategy, your team is already using AI at work.

They’re drafting emails with ChatGPT. Summarizing papers with Claude. Asking Copilot to clean up their spreadsheets. They’re not waiting for permission, and honestly, they’re not going to stop.

This type of “quiet pilot” isn’t a problem—it’s an opportunity. But only if you know what’s happening.

The risk of ignoring quiet pilots isn’t just governance (though that matters). It’s that you’re missing a real-time signal about where AI is already adding value in your organization. Your team is running experiments every day. Are you learning from them?

Make 2026 the year you bring these pilots into the light:

  • Ask the question. Survey your team—without judgment—about how they’re using AI tools today. You’ll be surprised by both the creativity and the frequency.

  • Look for patterns. Where are people finding value? What workflows are they augmenting? What problems are they solving that you didn’t even know existed?

  • Shape it thoughtfully. Establish appropriate guardrails where needed—particularly around confidential data, GxP activities, and external-facing content—but resist the urge to lock everything down. Overly restrictive policies just push usage underground.

There’s no putting the chatbot back in the box. The question is whether you’re learning from what your team is already doing—and channeling that energy toward meaningful opportunities.

6. Take Time to Define Your Roadmap

Not every company is ready for a major strategic planning initiative—and that’s okay. But as you head into 2026, it’s worth stepping back to acknowledge a simple truth: AI is here to stay.

The question isn’t whether to engage with it. The question is how deliberately you want to shape that engagement.

At minimum, AI tools will probably be useful to your team. People are already finding value in day-to-day tasks—drafting, summarizing, researching, analyzing. That alone justifies thoughtful adoption.

But the real opportunity lies in going further. As part of a well-considered roadmap, AI can return meaningful ROI for the business: accelerating timelines, derisking decisions, and unlocking capabilities that weren’t practical before. The difference between “useful tool” and “strategic asset” comes down to intentionality.

We’ve written before about how we think about building AI roadmaps for biotech companies—mapping opportunities to capabilities, prioritizing by impact, and sequencing implementation in a way that builds momentum without overwhelming the team.

You don’t have to do it all at once. But take the time to define where you’re headed. A little structure now will pay dividends as the technology—and your team’s fluency with it—continues to grow.

Kynetyk is a new type of partner for the biotech and life sciences industry: an Applied AI company focused on designing strategies that are accessible, practical, and effective for companies of any size. Reach out to us at hello@kynetyk.ai. We’d love to connect.