Dietitian Software Compared: Which Practice Platforms Have Useful AI?

Practice Better has the most useful AI for most dietitians, but it is an add-on: the AI Charting Assistant (dictation plus session summaries) starts at $10 a month per practitioner on Professional or above. Nutrium's AI notes work only on calls inside its own video tool, NutriAdmin sells AI recipes as an add-on, and Kalix's transcription is marked coming soon.

So the first question is not "which has AI?" but "where do I consult?". If most sessions run on Zoom or in a clinic room, Nutrium's notes will not help. If you see clients face to face, a dictation-and-summary tool helps more than one tied to video. And if your hours go into meal plans rather than notes, the note-taking AI is the wrong thing to pay for.

Follow me on Instagram@sagnikteaches

Where a dietitian's unpaid hours actually go

A first nutrition assessment gathers a diet history or recall, recent blood results, measurements, energy and protein needs, barriers to change and agreed goals. Afterwards the dietitian writes it up, often in ADIME format with a nutrition diagnosis written as a PES statement (problem, aetiology, signs and symptoms). Then comes the meal plan or handout, which can take longer than the note.

Connect on LinkedInSagnik Bhattacharya

That gives three separate admin jobs, and software helps each one differently:

Subscribe on YouTube@sagnikteaches

• Notes: an AI scribe or dictation tool turns the conversation into a draft note. This is the most mature AI on offer.
• Meal plans and recipes: AI drafts recipes to a target. Useful for variety, risky for anything clinical, and always needs checking.
• Client follow-up: AI drafts check-in messages and summaries of agreed goals. Cheap to do with a general assistant if your platform does not offer it.

Before choosing a platform, keep a one-week log of which of the three takes your time. For help separating what to automate from what stays with you, see which dietitian tasks to automate and which stay human.

Practice Better, Nutrium, NutriAdmin and Kalix compared

Prices and features below are from each vendor's pricing page as read on 4 November 2026, in USD. Check them before you buy; this category changes plans often.

Practice Better is the strongest all-round choice, but read its pricing page closely. The AI Charting Assistant is not part of any plan price; it is an add-on available from Professional up, priced by minutes of recording. Professional also lifts the client limit from 10 to 300, which is the upgrade most growing practices need anyway. The page lists several other AI features in its comparison table, but the flattened table does not make clear which plans include them, so confirm in a trial.

Nutrium's AI notes are generated during calls on Nutrium's own video tool, with an automatic transcript. Its pricing page says consultations run over Zoom or Google Meet still work in Nutrium but without the AI notes. That is a hard limit for anyone who sees clients in person or on another video platform. Its displayed prices are introductory, so ask what you will pay from year two.

NutriAdmin is built around assessments, meal plans and recipe documents. Its AI recipe generator, powered by ChatGPT, is sold as an add-on rather than included, so price it before choosing NutriAdmin for that feature. It names no AI note-taking. Kalix is a full practice system with ADIME, SOAP and DAP note templates and claims tools, priced by practice volume; its only AI-like item, telehealth transcription and summarisation, sits on a plan marked coming soon.

Which platform wins for which kind of practice

The table above lists features; the choice depends on how you work. A quick decision guide:

Two short walk-throughs show the guide in use. A dietitian running eight-week group programmes online, with 40 clients and few long notes, gains little from a scribe; Practice Better's programme features and its Plus plan's unlimited clients matter more, and the AI add-on can wait until notes become the bottleneck. A clinic-based dietitian seeing 18 clients a week face to face, with detailed gut-health notes, is the opposite case: Nutrium's video-only notes would never fire, so the choice is between Practice Better with the AI add-on and keeping a cheaper system with a separate general-purpose scribe. In both cases the decision came from one fact about the practice, written down before any demo.

If you already pay for a general assistant, comparing it with a platform add-on is worthwhile; choosing ChatGPT or Claude as a freelance dietitian covers that side.

An ADIME note from a recorded session, and what to correct

An illustrative excerpt from a follow-up session with a client recently diagnosed with type 2 diabetes:

DIETITIAN: "Hello [first name], let's look at your recent results and a typical day. Your doctor's letter says fasting glucose 7.8 mmol/L and HbA1c 7.2%. What's breakfast usually?"
CLIENT: "A big bowl of instant oats with honey and a glass of orange juice. By mid-morning I'm shaky and starving, so I get a sweetened latte and a muffin at work. Lunch is a deli sandwich. And lately I get bad bloating and cramps about half an hour after drinking milk."
DIETITIAN: "The shakiness fits a quick rise and fall after a breakfast that's mostly fast carbohydrate. The bloating after milk may point to lactose intolerance, which we can test with a short trial. Three goals: swap to rolled oats with chia and an unsweetened fortified plant milk; add about twenty grams of protein at breakfast, eggs or a protein powder; and keep a four-day food and symptom diary."

An illustrative draft note from an AI charting tool:

ASSESSMENT
- Labs: fasting glucose 7.8 mmol/L; HbA1c 7.2%.
- Intake: high-glycaemic breakfast (instant oats, honey, juice); mid-morning shakiness and hunger; sweetened latte and muffin; deli sandwich at lunch.
- GI: bloating and cramping after cow's milk; suspected lactose intolerance.

DIAGNOSIS (PES)
- Excessive carbohydrate intake related to frequent high-glycaemic meals and sweetened drinks, as evidenced by HbA1c 7.2% and mid-morning shakiness.
- Food- and nutrition-related knowledge deficit related to no prior education on carbohydrate distribution.

INTERVENTION
- Nutrition prescription: 1,800 kcal, consistent carbohydrate of 40-45 g per main meal; 14-day lactose elimination trial.
- Education: rolled oats, chia, unsweetened fortified plant milk; 20 g protein at breakfast.

MONITORING AND EVALUATION
- Review 4-day food and symptom diary at follow-up in three weeks.

Most of it is right, and it took seconds. Now look at what nobody said: the 1,800 kcal prescription, the 40-45 g carbohydrate target, the 14-day length of the lactose trial and the three-week follow-up. Each is a reasonable-sounding number the tool filled in. If you sign the note, they become your prescription. The fix is a review habit: read the intervention section against your own memory of the session, delete every figure you did not set, and add the ones you did. Also check the PES statements against the terminology your profession uses, because tools paraphrase them. For prompts and templates for this part of the job, see drafting ADIME notes with AI.

A solo dietitian's monthly sum with Practice Better and the AI add-on

An illustration with assumed figures. Picture a solo dietitian with 28 active clients and 14 sessions a week, averaging 45 minutes, currently on a $25-a-month scheduling app with notes typed into document templates. Her weekly admin log shows about 15 minutes of notes per session (3.5 hours), 4.5 hours on meal plans and 4 hours of messages and intake review.

She moves to Practice Better Professional at $59 a month billed annually, for the client portal, programmes and charting templates, and adds the AI Charting Assistant. Fourteen 45-minute sessions a week is about 2,730 minutes of recording a month (14 x 45 x 4.33), which is over the 1,200-minute tier, so she needs the $20 tier with 3,600 minutes.

At $110 a follow-up session, the extra $54 is covered by half a session a month. If the note checking really takes five minutes, she gets back about ten hours a month, which she can use for two or three more clients a week or for evenings off. The figure to test in the trial is the five minutes: if the drafts need heavy editing, as in the example above, it may be eight or ten, and the saving halves. Notice too that meal plans are her biggest block of time and this stack does nothing for them yet; that is the next thing to test, with a recipe add-on or a general assistant. The tutorial on AI tools for dietitian admin and client follow-up covers the options.

Data protection questions for nutrition platforms

Diet histories often contain eating-disorder history, gut conditions, medications and mental-health details. They do not belong in spreadsheets or consumer apps, and a platform that records sessions needs the same scrutiny as any clinical system. Ask every vendor three questions in writing:

1. Agreement: will you sign a data processing agreement (or the equivalent your rules require), and does it say client records are encrypted and not used to train AI models?
2. Audio and transcripts: how long are recordings and transcripts kept, can I delete them myself, and can a default change without notice? Vendors do change privacy defaults, so check the settings again at each renewal.
3. Export: can I export client profiles, notes and invoices in CSV or PDF in bulk if I leave? Test it in the trial with a dummy client.

Then tell clients. A line in your intake pack and a spoken check before you start recording ("I use a tool that drafts my notes from our conversation; is that all right?") is the minimum. Clients who decline get typed notes and no comment.

Insurance billing: making notes support the claim

Where nutrition therapy is billed to insurers, the codes are usually time-based and split into initial assessment, reassessment and group sessions. Claims get rejected when the note does not show the time, the setting and what was done. AI helps here only if it records those facts accurately; it cannot invent them.

A useful habit is a closing sentence spoken at the end of each session for the tool to capture: "Follow-up individual session, 40 minutes face to face, reviewed food diary and glucose readings, adjusted breakfast plan." Kalix, for example, lists claims tools and a managed clearinghouse add-on at $15 a month per billing provider, but the software can only check a note that a person wrote or approved. If a code suggestion appears, your billing contact confirms it against the payer's current rules.

Five ways dietitians use AI tools, and what goes wrong

1. Allergy-safe recipe drafts

Ask a general model for "a gluten-free dinner plan" and it may include ordinary soy sauce, malt vinegar or oats without saying they must be certified gluten-free. Whether you use NutriAdmin's AI Recipes or a general assistant, write the exclusions as hard rules:

Client: adult with coeliac disease and an anaphylactic peanut allergy.
Task: 5 dinner recipes, about 550 kcal and 35 g protein each.
Hard exclusions:
1. No wheat, barley, rye, spelt, triticale or oats unless marked certified gluten-free.
2. Name tamari instead of soy sauce; no malt vinegar; stock cubes only if labelled gluten-free.
3. No peanuts, peanut oil or peanut-derived ingredients. Flag any packaged ingredient with a "may contain" risk.
4. Under each recipe, add an "Allergen check" line listing every substitution made.

Illustrative output line: "Allergen check: tamari (gluten-free) replaces soy sauce; gluten-free stock cube; no nuts used." What to fix: the model cannot read labels, so anything packaged still needs you or the client to check the actual product.

2. Estimating a meal photo

Clients often send plate photos instead of diaries. A general assistant can estimate a plate (with no client name attached), and an illustrative estimate looks like this:

Estimated plate:
- Grilled chicken breast, about 140 g: 230 kcal, 43 g protein
- Cooked quinoa, about 150 g: 180 kcal, 6 g protein, 33 g carbohydrate
- Roasted broccoli with oil, about 120 g: 75 kcal
Total: about 485 kcal, 52 g protein
Uncertain: the chicken looks glazed; a sweet marinade would add sugar not counted here.

Use it as a conversation prompt ("was that glaze honey?"), not as data. Portion guesses from photos can be well out, and a diary built from them gives false precision.

3. A check-in message between sessions

Prompt: "Write a 75-word check-in for [first name], in week two of a low-FODMAP elimination phase for IBS. Acknowledge the cooking effort, ask whether bloating has settled, remind them this phase is temporary and ask them to message confusing labels through the portal." Illustrative output:

Hi [first name], two weeks in, well done: the elimination phase takes real planning and cooking. How has the bloating been this week? Remember this stage is temporary; we will start bringing foods back soon, one at a time. If you find a label you are unsure about, snap it and send it through the portal and I will reply within a day. [your name]

What to fix: "within a day" is a promise; keep it only if it is true. Check the phase length matches what you actually agreed.

4. Micronutrient gaps in an AI-drafted plant-based plan

AI meal drafts often hit protein, carbohydrate and fat targets while missing micronutrients. An illustrative audit of a draft plan for an adult woman, using reference values chosen for this example (use the reference intakes that apply to your clients):

The draft looked balanced on macros and would have failed the client on three nutrients. That is the case for keeping AI on drafting duty and you on sign-off.

5. Summarising a long intake form

New clients sometimes send eight pages: history, family history, sleep, stress and a long supplement list. Practice Better's comparison table lists an AI Session Prep feature; check whether your plan includes it. Otherwise a general assistant can summarise de-identified intake text. An illustrative summary:

PRE-SESSION SUMMARY
- Main concern: fatigue, brain fog after meals, 6 kg weight gain over 12 months.
- Conditions: Hashimoto's thyroiditis (on levothyroxine); mild iron-deficiency anaemia.
- Timing flag: takes iron supplement at the same time as levothyroxine. Check spacing with the client's prescriber or pharmacist.
- Pattern: skips breakfast; relies on convenience snacks; low fluid intake.

The timing flag is the valuable line, and it is phrased correctly: a prompt to check with the prescriber, not advice. If a summary ever tells you what dose to change, treat that as a fault in the prompt.

Five steps to switch platforms without losing records

1. Price your real stack. Add up what you pay now for video, booking, payments and note templates. Compare it with the new platform plus any AI add-on at the tier your recording minutes need, not the cheapest tier.
2. Build templates before migrating clients. Set up ADIME or SOAP templates with your PES fields and measurements in the new system first.
3. Trial the AI on mock sessions. Record two practice sessions with a colleague. Count what the draft got wrong, especially numbers you never said.
4. Set a meal-plan rule. No AI-drafted plan leaves the practice without your check of micronutrients, allergens and practicality.
5. Write team rules if you have staff. Who may record, what may be pasted into general assistants, and who signs notes. The tutorial on training a dietetic practice team on AI gives a starting set.

After a month, compare your admin log with the one you kept before. If note time has fallen and nothing wrong has reached a client record, keep the add-on. If you are rewriting most drafts, drop to the plan without it and use the savings elsewhere.

Dietitian practice software and AI questions

Can AI software write a clinically sound meal plan on its own?

No. It can draft recipes and days that hit calorie and macronutrient targets quickly, but it misses micronutrient gaps, hidden allergens, medication interactions and whether the client will actually cook it. Use the draft as a starting point and check every plan yourself before it reaches a client.

How is a dietitian AI note different from a general medical scribe?

A general scribe usually formats a SOAP note. Nutrition platforms let you map the summary into your own templates, including ADIME (assessment, diagnosis, intervention, monitoring and evaluation), so dietary recall, measurements and goals land in the right sections. Either way, you check every line, because scribes add details nobody said.

Do these platforms keep my consultation audio?

The pricing pages read for this tutorial do not say, so ask each vendor in writing: how long audio and transcripts are kept, whether either is used to train models, and how you delete them. Keep the reply with your contract and check it again at renewal, because vendors change defaults.

Which platform suits a solo dietitian starting out?

Practice Better's free Sprout plan (3 clients) or Starter ($25 a month billed annually, 10 clients) covers the basics, but its AI Charting Assistant needs Professional or above. Nutrium suits you if you consult on its own video tool. If meal-plan documents are your main product, look at NutriAdmin and its AI Recipes add-on.

Further reads

• Is an AI Scribe Worth It for a Small Private Clinic? — Whether a scribe pays for itself in a small private practice.
• AI Scribe Costs Compared: What Small Clinics Pay in 2026 — Scribe prices across small clinics, for comparison with add-ons.
• How to Roll Out an AI Scribe Without Losing Patient Trust — Introducing recording to clients without losing their trust.
• How a Small Clinic Can Use AI at the Front Desk Safely — Automating booking and enquiry replies around the consultations.
• How a Clinic Can Use AI to Cut Missed Appointments — Reminder set-ups that cut no-shows and late cancellations.
• What Is an AI Scribe and How Does It Work in a Consultation? — What a scribe does with the audio during a session.
• AI Tools for Nutritionist Practices: Client Check-Ins and Handouts — How a solo nutritionist can use AI for check-in messages, diary summaries and handouts, with review rules so no unsafe advice reaches a client.
• Nutritionist AI Prompt: Write Consent and Cancellation Wording — A fill-in-the-blanks prompt that drafts consent and cancellation terms for a nutrition practice, plus the checks before a solicitor sees it.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Practice Better pricing page; Nutrium pricing page; NutriAdmin pricing page; Kalix pricing page. All read 4 November 2026.