Kristin Gleitsman

Leadership in Biotech

Tag: tools

Illustration of a desk with a figure from a paper on ChatNT

ChatNT: The future of biological assistants—or a mirage in a lab coat?

By

On June 17, 2025

In AI ∩ Bio

The team behind ChatNT introduces a conversational AI agent trained to perform 27 genomics, transcriptomics, and proteomics tasks—by prompting it in plain English. Built on a DNA encoder (Nucleotide Transformer v2) and a frozen English decoder (Vicuna-7B), ChatNT achieves state-of-the-art or near-parity performance with many specialized models, solving tasks like splice site detection, RNA degradation prediction, and protein melting point estimation—all through natural language queries.

But before we celebrate too loudly…

What does it mean when we start predicting complex molecular properties by chatting with a model—and trusting the answer without understanding the underlying biology? ChatNT lowers the barrier to entry, making powerful models accessible to those without deep bioinformatics expertise. That’s a design strength—but also a risk. Scientific depth, if not deliberately preserved, can quietly erode. We could end up with users who can write reasonable prompts but lack the scientific grounding to recognize when the answers are wrong or incomplete, and don’t have the foundational knowledge needed for scientific creativity.

To their credit, the authors do include a post hoc, perplexity-based calibration method to understand the model’s confidence in its answer (in other words, they check how confidently the model would have chosen its answer by measuring how surprised it is by different options after the fact). But there’s no real-time uncertainty alert, no embedded safeguard for when the model is operating outside its training distribution—just statistical proxies layered onto a system that still speaks with unwarranted certainty. In regulated or high-stakes domains like diagnostics, that’s absolutely not enough. Hallucinations don’t come with warning labels. And a well-attributed motif—say, a TATA box or splice site—is no guarantee of biological correctness.

From a diagnostics strategy perspective, ChatNT is a credible preview of what’s coming: a unified interface for interpreting multi-omics data and compressing complex workflows into a single prompt. But we are not there yet. Trust, fidelity, and epistemic transparency remain unsolved. For now, these models should be treated as useful but fallible junior collaborators—not autonomous copilot researchers in their own right.

To early-career scientists: this is your edge. Tools like ChatNT are remarkable—but only in the hands of those who still understand the biology. The future still belongs to those who can spot an implausible claim, who know how to interrogate things from first principles, and who can still deploy their own knowledge to connect disparate dots and generate novel scientific hypotheses. Your role isn’t to step aside. It’s to double down on understanding, so that you can interrogate, shape, and lead the evolution of these tools.

AI ∩ Bio

Illustration of scientific paper with the title "AI ∩ Bio: Reading the Revolution"

How I’d want my team to use this tool: Use ChatNT to validate hypotheses you’ve already reasoned through—not to generate them in isolation. Let it help challenge assumptions, spot inconsistencies, or simulate mechanistic alternatives based on sequence features. Think of it as a fast, articulate assistant: useful for in-silico hypothesis exploration, not for making experimental decisions without expert oversight. And never input PHI or proprietary data into public-facing AI tools. Period.

Image of a construction manager looking at an oversized tablet computer to select from one of three robots standing in front of them. In the distance, a city is under construction

Scaling Well in 2025: AI Tools That Actually Help

By

On April 1, 2025

In Scaling Well

Scaling Smarter Starts with Designing for Resilience

The most successful use of AI in 2025 won’t necessarily come from technological breakthroughs. It will come from thoughtful rethinking how their organizations operate, reshaping how time, knowledge, and decisions flow across our organizations.

AI tools should support and augment the pillars of scaling that I’ve written about previously. They should free up capacity for deep work, enable culture that scales, and support infrastructure that can carry the weight of change.

This means the companies scaling well in 2025 aren’t asking, “How do we use AI?” They’re asking, “How do we use AI to make our company better, faster, and more resilient?”

In this piece, we explore how AI can support each of those pillars. Each section includes real-world tools, examples, and decision cues to help leaders think more clearly about what to pilot and how to scale well.

Let’s dive in.

Designing for Slack & Innovation

“Slack isn’t waste. It’s where creativity, resilience, and serendipity live.”

In the original article, we framed slack not as inefficiency, but as infrastructure for innovation—a concept grounded in queuing theory and visible in everything from engineering velocity to decision-making latency. In 2025, AI makes it possible to unlock “smart slack”: creating time, reducing bottlenecks, and diffusing knowledge—all without adding headcount.

Below are three AI tools that enable this.

Tool 1: Glean – Contextual Search to Reduce Bottlenecks

Problem: Teams lose time recreating knowledge that already exists—or waiting on the one person who has the answer.

Tool: Glean indexes all your internal systems (Docs, Slack, Jira, Drive, Confluence, etc.) and uses AI to deliver relevant, secure search results contextualized to the user.

Impact: Reduces time spent answering repeat questions, reduces key-person dependency, and reclaims hours per week for deep work.

Real Example: Confluent Case Study – Glean

Piloting Tip: Start with one department and track internal support volume, time-to-answer, and repeated questions. Pair with light documentation practices to amplify results.

How to Know if This Belongs on Your Roadmap

  • You’re hearing complaints about repetitive internal questions or knowledge buried in Slack
  • Onboarding timelines are dragging because institutional knowledge isn’t searchable
  • A few individuals are serving as knowledge routers across functions
Glean is worth piloting if you want to reduce dependency on key people and unlock capacity without increasing headcount. Ask your department leads how much time is spent answering questions that should already be documented.

Tool 2: Notion AI – Workflow Co-Pilot to Free Cognitive Slack

Problem: Knowledge workers are overloaded with routine documentation, synthesis, and follow-up work.

Tool: Notion AI automates meeting summaries, drafts project plans, and generates next steps across team workspaces.

Impact: Frees up “cognitive slack” for strategic thinking, synthesis, and creative problem-solving.

Real Example: Figma Case Study – Notion

Piloting Tip: Begin with retros, meeting notes, or OKR drafts. Then explore more structured workflows like customer issue triage or roadmap generation.

How to Know if This Belongs on Your Roadmap

  • Your OKR cycles feel slow or painful to prep
  • Meeting notes, follow-ups, and documentation tasks are draining your highest-leverage people
  • Strategic initiatives stall because everyone’s too busy with tactical maintenance
Notion AI is worth funding if your teams need more “thinking time” but can’t find it. Look for small pilots where cognitive load is high but output is low-value admin work.

Tool 3: Stack Overflow for Teams – Capture and Share Tribal Knowledge

Problem: Tribal knowledge lives in people’s heads or buried in Slack threads—causing bottlenecks and slow onboarding.

Tool: Stack Overflow for Teams creates a company-specific Q&A hub. AI enhances search, recommends related questions, and flags potential duplicates to reduce noise and improve speed-to-answer.

Impact: Accelerates onboarding, reduces repetitive internal questions, and builds resilience when key people leave.

Real Example: Box Case Study – Stack Overflow for Teams

Piloting Tip: Seed it with your most-asked questions. Create a recognition system for contributors. Link key answers directly from onboarding flows.

How to Know if This Belongs on Your Roadmap

  • You’re hearing “I didn’t know that existed” across teams
  • Ramping new hires feels fragile and overly dependent on 1:1 support
  • The same technical questions are being asked—and answered—on repeat
This tool is worth considering if your resilience is too dependent on institutional memory. Ask your leads how much time their teams spend re-answering solved problems.

Operationalizing Slack

Strategic slack isn’t found. It’s designed. The above tools create visibility into where time, attention, and knowledge are leaking—but the next step is turning that visibility into action.

  • Map your bottlenecks: Where does work pile up or depend on a few people? Use these tools to surface hotspots.
  • Reclaim 10–20% capacity: Track time saved, then explicitly reinvest it into innovation sprints, mentorship, or deep work time.
  • Start with one team: Pilot in a high-friction area with motivated adopters. Share wins before scaling org-wide.
  • Don’t over-automate: Tools should augment, not replace, human judgment. Start simple, layer complexity later.

Reference: For deeper guidance on build vs. buy vs. fine-tune, see Appendix A.

Once you’ve carved out slack, the next challenge is ensuring your teams use it well—without needing top-down oversight every time. That starts with culture.

Scaling Culture & Leadership

“Culture is how decisions get made when no one is watching.”

If slack gives teams processing power, then culture is the operating system that determines how that power gets used. It determines how teams behave under pressure, how decisions get made across time zones, and how mission translates into action when leaders aren’t in the room.

In 2025, generative AI unlocks new ways to scale cultural alignment, leadership development, and human connection—without requiring armies of coaches or L&D staff. Below are three tools that make culture more scalable, distributed, and actionable.

Tool 1: Humu – Culture-Driven Behavioral Nudges at Scale

Problem: Culture often degrades in the middle. Execs may model values, but middle managers struggle to reinforce them day-to-day.

Tool: Humu delivers science-backed, personalized nudges to encourage behaviors aligned with company goals.

Impact: Reinforces key behaviors across roles, time zones, and contexts—without requiring more oversight.

Real Example: Humu Case Studies

Piloting Tip: Start with one leadership behavior to reinforce—like recognition, inclusive meetings, or timely feedback. Layer nudges on top of any existing training or communications push.

How to Know if This Belongs on Your Roadmap

  • You’re struggling to turn stated values into visible day-to-day behaviors
  • Cultural consistency breaks down between levels or locations
  • Your engagement surveys surface issues like low feedback or recognition
Humu can bridge the gap between intention and action. Prioritize it if you’re investing in manager enablement or aiming to reinforce cultural behaviors without scaling your HR team.

Tool 2: Pinnacle AI – Custom Coaching & Leadership Simulations

Problem: Leadership coaching rarely scales beyond the executive level.

Tool: Pinnacle AI enables companies to create custom GPT-based avatars that simulate high-stakes conversations, coach leaders in real time, and reinforce values.

Impact: Expands coaching access, builds soft skills through practice, and embeds culture into day-to-day behavior.

Real Example: Pinnacle AI Overview

Piloting Tip: Start with a common friction point (e.g., giving tough feedback). Use a leadership avatar to simulate the conversation, then track usage and impact.

How to Know if This Belongs on Your Roadmap

  • You need to scale coaching or practice-based leadership development without adding headcount
  • Managers are hesitant or underprepared for hard conversations
  • Soft skill gaps are impacting team dynamics or delivery
Pinnacle is worth trying if you’re investing in middle manager capability-building or want to embed feedback culture through safe, scalable practice.

Tool 3: Together Platform – Peer Mentorship with AI-Enhanced Matching

Problem: Cultural transmission often relies on informal relationships that don’t scale.

Tool: Together uses AI to match employees into mentorship relationships based on goals, skills, and values.

Impact: Reinforces culture through peer connection. Supports belonging and cross-functional leadership.

Real Example: Together Case Studies

Piloting Tip: Start with onboarding cohorts or ERGs. Use AI-generated prompts to anchor discussions in real cultural moments and shared learning goals.

How to Know if This Belongs on Your Roadmap

  • You’re looking to scale belonging and connection in a hybrid or distributed org
  • Current mentorship programs are ad hoc or unsustainable
  • Peer learning feels like a missed opportunity across levels or silos
Together is a strong fit if you’re building internal talent pipelines or trying to support new hires and ERGs without overloading HR.

Cultural Playbook: Scaling Values, Not Just Policies

  • Pick your pressure points: Identify where cultural gaps show up—new manager transitions, post-reorg teams, or moments of change. Deploy nudges (Humu), simulations (Pinnacle), or mentorship (Together) at those friction points.
  • Time it right: Behavior change sticks when it’s timely. Align tools with onboarding, performance reviews, or team formation.
  • Reinforce with data: These tools generate signals. Use them to monitor alignment, not just activity.
  • Connect the dots: Use Together to reinforce Pinnacle simulations. Follow Humu nudges with live discussion prompts.

Of course, with every new AI tool comes another integration, another dashboard, another cognitive switch. It’s worth asking: do we really need three tools, or can one flexible platform get us 80% of the way there? This isn’t just a user experience issue—it’s an infrastructure one. And that’s where we go next.

The Case for Fewer, Flexible Tools

Leaders face a strategic question: should you prioritize tools that do one thing exceptionally well, or ones that solve multiple problems “well enough”?

Some considerations:

  • If your org is resource-constrained, favor tools that integrate natively with your existing stack (like Pinnacle AI into Slack) and solve for multiple moments (onboarding, feedback, DEI).
  • If your org is scaling fast with complex silos, best-in-class tools may still make sense—as long as IT can support them.

Several tools featured here offer multi-purpose value: Pinnacle AI can support onboarding, leadership development, and cultural alignment. Together can double as an onboarding accelerator and a peer learning platform. We’ll explore how to evaluate tools across breadth vs. depth in Appendix B.

The Hardest Part Isn’t the Tool. It’s the Terrain.

Cross-functional adoption is where AI efforts live or die.

Even the best tool won’t gain traction if:

  • IT says they’re already at capacity
  • Legal worries about risk
  • Enablement teams are juggling five other rollouts
  • Or end-users don’t trust the tool (or the motive)

If this sounds familiar, try this instead:

  • Start where there’s pull. Pilot with teams who are already stretched and asking for help.
  • Solve their problem, not yours. Frame the tool around a pain they already name.
  • Invite, don’t mandate. Use early adopters to tell the story, not top-down pressure.
  • Build in feedback. Create fast feedback loops between teams and tool sponsors.

Adoption isn’t just change management—it’s coalition-building. And how you roll out the first tool will shape whether people lean in or tune out when the next one comes.

Infrastructure & Enablement: Lightening the Load, Quietly Powering Scale

“You can’t scale what you can’t support.”

Slack and culture shape what’s visible—but infrastructure is the foundation. It decides how much complexity your organization can carry, and how well it holds up under change. And yet, infrastructure teams—especially IT, platform, and internal tooling—are often the most overburdened and underappreciated.

Generative AI won’t solve infrastructure debt overnight. But it can reduce manual load, increase visibility, and turn reactive work into proactive systems management. Below are a few tools helping infrastructure teams do more with less—and sometimes even breathe.

Tool 1: Moveworks – Conversational AI for IT Support

Problem: IT teams are flooded with repetitive tickets, onboarding requests, and access issues.

Tool: Moveworks resolves common IT support requests instantly through Slack, Teams, or email.

Impact: Reduces ticket volume, accelerates onboarding, and frees up IT capacity for more strategic work.

Real Example: Moveworks Case Studies

Piloting Tip: Start with high-volume, low-complexity tickets. Track resolution speed, team bandwidth, and internal satisfaction.

How to Know if This Belongs on Your Roadmap

  • Your IT support teams are drowning in high-volume, low-complexity tickets
  • Employees complain about slow or inconsistent onboarding and access
  • You want to improve internal experience without increasing IT headcount
Moveworks is worth piloting if you want to unlock IT capacity for infrastructure and security work. Ask: where could automation improve internal credibility?

Tool 2: Cortex – Internal Service Catalogs & Engineering Visibility

Problem: Engineering teams lose time navigating unclear ownership and shadow systems.

Tool: Cortex offers internal scorecards, visibility tools, and service catalogs that clarify what’s running—and who’s responsible.

Impact: Supports platform health, reduces hero culture, and increases resilience.

Real Example: Cortex Customers

Piloting Tip: Map key services with unclear ownership. Use Cortex to build scorecards and reinforce accountability.

How to Know if This Belongs on Your Roadmap

  • You’re seeing repeated outages or friction due to unclear service ownership
  • Engineering productivity depends on a few heroic individuals
  • Your team lacks visibility into internal reliability or SLA gaps
Cortex is a fit if you’re trying to mature your platform engineering or prepare for scale. Ask: who owns what—and do they know it?

Infrastructure Insight: Design for Longevity, Not Just Load

“If you’re thinking about infrastructure only when it fails, you’re already behind.”

  • Engage infrastructure early when evaluating AI tools—especially those touching data or permissions.
  • Give infrastructure leaders a seat at the table in AI strategy and tooling decisions.
  • Track infrastructure health visibly—and treat internal platforms as core products.
  • Recognize and reward the teams that make scalable systems possible.

Conclusion: Scaling Well Means Leading Differently

The leaders pulling ahead aren’t just adopting tools—they’re rethinking how their organizations operate. They’re treating slack as strategy, culture as infrastructure, and infrastructure as leverage.

This transformation isn’t really about the tech—it’s a business transformation that demands real leadership and coordination.As KPMG’s Vice Chair of Artificial Intelligence (AI) and Digital Innovation, Steve Chase, put it, we’ve entered a phase where companies need someone to lead the movement—someone who can cross silos, hold tension, and help the organization build real muscle memory around transformation.

Leading the movement to implement AI successfully means starting with intention—with intelligent pilots, cross-functional alignment, and a clear view of what you’re solving for.

The tools are here.

The stakes are real.

And how you lead this transition will define how well your company scales… or stalls.

Appendix A: Choosing the Right Tool – A Strategic Framework

Before choosing a tool—or even beginning a pilot—executives should step back and ask: What kind of leverage are we trying to create, and what constraints must it navigate?

This framework provides a higher-level lens to help you prioritize the right kind of AI investment:

1. Clarify the Friction

What are you actually trying to solve?

  • Time bottlenecks: Teams are stuck doing repetitive work or can’t move fast enough
  • Cultural inconsistency: Values and behaviors vary wildly across functions or levels
  • Infrastructure drag: Tooling, access, or service ownership are slowing everything down

2. Map Organizational Readiness

How prepared is your org to absorb a new AI tool?

  • Technical readiness: Do you have IT capacity and integration capability?
  • Behavioral readiness: Will people use it—or resist it?
  • Strategic alignment: Does this tool support a broader goal you already care about?

3. Decide the Mode of Leverage

What kind of support are you trying to amplify?

  • Automate: Remove low-leverage tasks (e.g. ticket triage, note-taking)
  • Augment: Enhance high-leverage talent (e.g. coaching, mentoring, cross-functional insight)
  • Diffuse: Spread knowledge and behavior (e.g. onboarding, Q&A, nudges)

4. Start Where You Have Urgency + Ownership

Don’t start where the problem is loudest—start where a function wants the help.

  • Look for “pull,” not “push” opportunities
  • Pilot in functions already experimenting or stretched thin but open-minded

This framework helps you move from “interesting tool” to “clear decision.”

Example: Applying the Framework by Company Size

Here’s how this framework might guide tool selection across company stages, using high-leverage, multi-purpose tools that can solve for multiple frictions:

Small Company (≤50 employees)

  • Friction: Time bottlenecks, no L&D team, documentation debt
  • Readiness: Agile culture, limited IT bandwidth
  • Mode of leverage: Automate and diffuse
  • Tool Match: Notion AI for streamlining workflow + basic coaching content; Together for onboarding + connection-building

Medium Company (50–300 employees)

  • Friction: Scaling teams, mixed leadership capability, mounting ticket load on IT
  • Readiness: Willing early adopters in some functions
  • Mode of leverage: Diffuse and augment
  • Tool Match: Pinnacle AI for scalable leadership coaching + feedback support; Moveworks to relieve IT and improve employee experience

Large Company (300+ employees or multiple sites)

  • Friction: Cultural inconsistency, distributed knowledge, infrastructure complexity
  • Readiness: Dedicated IT, HR, and Enablement functions—but political complexity
  • Mode of leverage: Systematize and augment
  • Tool Match: Pinnacle AI for global leadership training + cultural reinforcement; Notion AI to reduce admin burden; Moveworks for automated support across functions

→ Use this framework not to pick “the best tool,” but to clarify what kind of leverage your org actually needs.

Appendix B: Questions to Evaluate AI Tools Before You Commit

Before piloting or purchasing an AI tool, ask:

Strategic Fit

  • What exact friction is this solving?
  • Will it augment high-leverage work—or just automate noise?

Integration & Infrastructure

  • Does it work with our current stack (Slack, Teams, Drive, etc.)?
  • How much IT lift is required?
  • Who owns implementation?

Cultural & Change Readiness

  • Will people trust and use this tool?
  • Does this require behavior change—and do we have a plan for that?

Measurement & Momentum

  • What leading indicators will show this is working?
  • What story will we tell if we want to scale it org-wide?

Example: Evaluating Two Tools in Practice

Let’s say you’re weighing Notion AI vs. Pinnacle AI—both multi-use tools with clear executive appeal, but very different strengths.

Scenario 1: Notion AI

  • Strategic Fit: Your teams are overloaded with repeat documentation and task spillover.
  • Integration: Already using Notion lightly; integration is easy and user-friendly.
  • Cultural Fit: Teams are open to automation, especially if it feels like a personal productivity win.
  • Measurement: Track adoption in meetings and goal-setting workflows; estimate time saved weekly.
Great pilot if your goal is reclaiming capacity and improving org hygiene. Low-risk, broad benefit.

Scenario 2: Pinnacle AI

  • Strategic Fit: You’re struggling to scale coaching, leadership readiness, and feedback culture.
  • Integration: Can be deployed via Slack or Teams, with configuration support available.
  • Cultural Fit: Managers want help but don’t have time for traditional development.
  • Measurement: Track usage of leadership avatars, user satisfaction, and reported confidence in high-stakes conversations.
Worth piloting if you’re focused on middle-manager enablement or cultural consistency. Medium complexity, but potentially high cultural leverage.

These tools don’t compete—they complement. But knowing what type of leverage you need first is key. Use these questions to sharpen internal alignment, clarify trade-offs, and set realistic success metrics before making big bets.

Powered by WordPress & Theme by Anders Norén