How AI Tools Are Reshaping Workflows for Online Coaches

In the evolving landscape of strength and conditioning, large language models (LLMs) and custom-built AI tools are transforming how coaches program and support their clients. Rather than replacing the coach, these tools are increasingly serving as powerful assistants: automating administrative tasks, preserving contextual details, and giving coaches more bandwidth to focus on the creative and relational aspects of their craft.

From Templates to Personalized Programming

Most coaching platforms allow for templated programs that can be customized with starting weights and scheduled throughout a training cycle. This can be helpful, but it still requires coaches to manually make nuanced adjustments and remember client-specific details. Recent integrations with LLMs have advanced this process.

By feeding previous training data and coach–client interactions into a custom AI agent, coaches can generate fully personalized 8–16 week programs that incorporate appropriate loading, exercise selection, scheduling, and even personalized notes or cues. Unlike static templates, the LLM can pull from the client’s training history and feedback to embed relevant cues automatically into future exercises.

For example, if a coach consistently cues a client to “brace earlier” in their deadlifts, that note can be carried forward automatically into subsequent training blocks. This level of detail is often lost in manual programming, simply because coaches are managing dozens of athletes and do not have the bandwidth to comb through every past note.

Integrating Coaching Feedback and Transcripts

Another powerful development is the ability to pull video feedback transcripts directly into the context window of the AI agent. Many clients have expressed a desire to see cues from their last feedback video during their next workout. With transcript integration, this becomes automatic. The AI can distill key coaching points from previous videos and embed them into the programming notes.

This does not replace coaching. It augments it. Coaches can continue to deliver high-value asynchronous video feedback, while the AI ensures that none of that information gets lost between training sessions.

Building Better Tools With Minimal Overhead

This technology is not limited to programming logic. Coaches and developers have built lightweight tools to unify and query client data more effectively. For example, if all client comments and messages are aggregated in chronological order, it is easy to search for issues like, “When did shoulder pain first come up?” It is also easy to make quick estimations of one-rep maxes to assess recent performance trends without manually digging through old PR data.

Surprisingly, much of this development can be done with minimal cost. Running a Grok or GPT-5 integration to generate six-week programs, for example, may only cost a few dollars. And for many coaches, the tools are simple enough to adapt without formal software training. LLMs can generate the base code for new features, and coaches can refine them iteratively.

Automating the Grunt Work, Elevating the Coach’s Role

A useful analogy for AI’s role comes from the medical field: a highly skilled heart surgeon does not prep the patient or draw blood. Support staff handle those tasks so the surgeon can focus on what only they can do. Similarly, AI tools can handle repetitive, administrative, or detail-oriented tasks that traditionally consume a coach’s time—tracking cues, formatting programs, calculating loads—freeing the coach to focus on creativity, decision-making, and client relationships.

For example, AI can summarize client’s training history into a detailed month-by-month report highlighting PRs, setbacks, and progress trends. This kind of analysis would take hours to assemble manually, but AI can generate it in minutes, adding significant value for the client without increasing the coach’s workload.

Practical Development Lessons: Keep Context Small

One key technical insight for those building their own tools is that smaller context windows often work better. While LLMs now support increasingly large context lengths, simply dumping all historical data into the model leads to hallucinations and inefficiencies. Instead, the more effective approach has been to have the model generate summarized reference documentation from raw data (e.g., code bases, client histories) and feed those summaries into the context window as needed.

This modular approach not only keeps prompts cleaner but also improves accuracy and reliability, especially for program generation.

Open APIs and Custom Tools for Non-Developers

For coaches who want to experiment with building their own solutions, Karl Schudt—a coach and developer—has made a free, open-source toolkit available via GitHub. This includes:

• A unified feed of comments and messages
• Downloadable workout histories
• Automated program generation using Grok or GPT models
• Customizable context windows for programming style preferences
• Searchable data repositories for client-specific context (e.g., equipment lists, goals, lab results)

These tools can be forked and adapted, allowing coaches without formal development backgrounds to build custom solutions. With an LLM’s help, even those new to coding can follow step-by-step guidance to set up their own versions to pull and manipulate client data, generate programs, or analyze performance trends.

Incremental Adoption and Long-Term Gains

Perhaps the most important lesson Karl shared was not to expect linear progress with your experiments. Adopting AI tools can initially cause a dip in productivity as you learn to work with new workflows and wrestle with integration challenges. But over time, coaches who persist often experience significant gains in efficiency and quality, much like any major process improvement attempt.

Starting small—such as generating simple scripts, reports, or programming blocks—allows coaches to build familiarity and confidence without overhauling their systems overnight.

AI tools are not replacing coaches. They are extending their capabilities. By offloading administrative and repetitive programming work, coaches can devote more energy to what matters most: applying their expertise, building relationships, and delivering exceptional coaching experiences at scale.

Explore Karl Schudt’s Open-Source Tools

If you’re ready to experiment with AI-driven workflows, check out Karl Schudt’s free GitHub toolkit for online coaches. It includes automated program generation, searchable client data, and customizable templates you can adapt to your own coaching systems.

This material was recently covered in the Business of Coaching Workshop, a series designed to help coaches grow their businesses by mastering key principles like trust, pricing, and delivering value. Each session dives into actionable strategies to build better client relationships and drive success. Want to take your coaching practice to the next level? Join us for the next workshop—it’s free.