Personalized treatment plans take time. You research diet, lifestyle, and movement recommendations for each patient, then adapt them for restrictions, age, and clinical priority. Most practitioners spend longer on this work than they'd like.
Recommended Actions is a new feature inside Optimal DX's Treatment Plan Builder. When you've selected your patient's health concerns from their blood test results, you click Generate Recommendations, and the AI drafts personalized diet, lifestyle, and movement recommendations based on those concerns and the patient's preferences. The recommendations appear in an editor where you can rewrite them, add to them, remove what doesn't fit, or replace anything entirely. Nothing reaches the patient until you've reviewed and approved it.
How it fits your workflow
The feature lives where you'd expect it: inside the existing Treatment Plan Builder, as Step 3. The workflow you already know stays the same. You select your patient, choose the blood test, review the health concerns identified by the Functional Health Report, and add supplements. Recommended Actions then drafts the diet, lifestyle, and movement layer for you, all in one pass, tied to the health concerns you've selected. (For the full picture of how Treatment Plan Builder works, see the Treatment Builder page.)
The personalization layer
The system covers 78 health concerns, with content drawn from the Functional Medicine and Functional Blood Chemistry knowledge base our research team has been building for years. It weighs concerns by likelihood, so higher-likelihood concerns get more attention in the recommendations. It resolves conflicts when one recommendation contradicts another, choosing the safer option. It adapts to patient age, with gentler recommendations for older adults.
And it accounts for patient preferences. Twenty different preference options cover dietary restrictions (vegan, gluten-free, diabetic), physical limitations (restricted mobility, chronic pain), and lifestyle factors (night shift, limited time, no kitchen access). The preferences are set in the patient profile once, and Recommended Actions pulls them in automatically every time you generate a treatment plan for that patient. A 70-year-old vegan patient with restricted mobility gets recommendations that respect all three, every time.
The AI draws on that vetted knowledge base. It's a meaningfully different starting point than the open internet, and it's what makes responsible AI possible in our context. The recommendations read like Functional Medicine because the source material is Functional Medicine.
A practical example
Say your patient has high cholesterol indicated by their blood test, alongside markers suggesting blood sugar dysregulation and possible thyroid imbalance. They're 64, vegan, and have restricted mobility from a knee replacement. In the old workflow, you'd research diet recommendations for each concern, cross-check them against the vegan restriction, adapt the movement recommendations for limited mobility, and prioritize the whole thing by clinical urgency. Twenty minutes if you're fast, thirty if you're being thorough.
With Recommended Actions, you select the three health concerns, the patient's profile pulls in their preferences, and you click Generate. The system drafts diet recommendations that respect the vegan restriction, movement recommendations that account for the knee, and it weighs the three concerns against each other, prioritizing accordingly. You review what comes back, edit anything that doesn't fit your judgment, and move on.
The practitioner stays in the lead
I wrote about how we think about AI in a recent founder's note. The short version: AI drafts, practitioners decide. Recommended Actions follows that principle to the letter. Every recommendation the AI generates lands in an editor for your review. You can keep what works, rewrite what doesn't, add your own clinical judgment, or replace anything entirely. The AI gets you to a working draft. You make the final call before anything reaches the patient. We built the feature this way on purpose.
What this means in practice: Recommended Actions handles diet, lifestyle, and movement recommendations tied to each health concern, prioritizes by likelihood, resolves conflicts when they arise, and adapts to patient age and preferences. What stays with you is the clinical interpretation of the blood test results, the final approval of every recommendation, and any treatment decisions.
Available in Optimal DX
Recommended Actions is live in Optimal DX now. It handles the drafting work in seconds, keeping the personalization built in and the clinical control with you.
This is the first AI feature in Optimal DX, and the way we built it is the way we plan to build the rest of them. There will be more in the coming months, and each piece will follow the same principles: AI drafts, practitioners decide, patient data stays protected, and the practitioner-patient relationship stays at the center.
If you want the longer version of how we think about AI in Functional Medicine, it's in the founder's note.
CLICK HERE If you want to see what Recommended Actions looks like in practice.
Ready to see Recommended Actions for yourself? Head over to our pricing page to review our plan details.
