The patient portfolio: every signal, every year, every system

The product most people see when they look at Helose is the briefing. It’s a single screen that a functional/integrative medicine physician opens 90 seconds before walking into the room. Five tracks, one shared time axis, the labs and supplements and meds and encounters and wearables laid out so the doctor can scan it in under a minute.

That’s the demo. It’s not the product.

The product is the patient portfolio. The briefing is one view of it.

What the portfolio actually is

The portfolio is the longest-possible record we can assemble for a single patient, across every system the clinic already runs. For a typical FM clinic that means:

• Labs from Rupa, LabCorp, Quest, Genova, DUTCH, GI-MAP, every panel the patient has ever drawn
• Supplements from Fullscript, including the date a regimen started and the date it stopped
• Medications from Charm or Cerbo, including the prescriber’s notes and the discontinuation reason where it exists
• Encounter notes from Charm, Cerbo, Practice Better, Healthie, verbatim, not summarized
• Wearable data from Oura, day by day, aligned to everything else
• Chart history that arrived by fax, OCR’d, dated, tagged to the right patient

None of that is new data. The clinic already paid for all of it. It’s just been sitting in six different systems that don’t talk to each other, which means no one has ever seen it as one record.

Helose assembles it. That’s the whole job.

Why the time axis matters

Every signal in the portfolio is anchored to a date. That sounds obvious until you try to do it. Rupa’s lab dates are in one format; Charm’s encounter dates are in another; Fullscript’s supplement start dates are the order date, not the date the patient actually started taking the pill. Oura is timezone-shifted. Faxed records from the old PCP have handwritten dates that need to be transcribed.

When all of that lands on one canvas, things become visible that were not visible before.

A TSH of 4.2 from 18 months ago, sitting next to a Levothyroxine prescription that started 17 months ago, sitting next to a TSH of 1.8 from last month, is a fact. Not an interpretation. The portfolio puts those three data points on the same horizontal line, in order, and lets the physician draw the line between them.

The physician already knew how to read that pattern. What was missing was the assembly.

What we deliberately don’t do

Helose does not say “the patient responded to thyroid replacement.” Helose shows the TSH at month -18, the start date of the Rx, the TSH at month -1. The verb belongs to the doctor.

Helose does not say “supplement adherence appears poor.” Helose shows the Fullscript order dates, the encounter quote where the patient described their routine, and the timing of the last refill. The inference belongs to the doctor.

This is a design principle, not an accident. We’ve written a separate post about why we built it this way. The short version: physicians don’t trust tools that draw conclusions for them, and the legal posture of a tool that assembles facts is meaningfully different from the legal posture of a tool that interprets them.

What becomes possible on one canvas

Once every signal lives on one time axis, a few things stop being hard.

Trend lines stop being a thing the physician has to reconstruct in their head between two tabs. The Free T3 from the Genova panel and the Free T3 from the LabCorp panel show up on the same line, with the lab source labeled, so a discontinuity between two assays is visible rather than buried.

Reference-range deviations stop being a thing the physician has to compute mentally for every row. Each lab value carries a badge showing where it sits relative to the FM target range, not just the lab’s printed reference range, which was set for a different question.

Chart history stops being a thing the physician has to read end-to-end before a visit. The verbatim quote from the prior encounter that’s relevant to today’s complaint is pulled into the briefing, with a link back to the full note. The portfolio is the full thing; the briefing is the curated extract.

Transfer notes from a faxed chart, six months ago, stop being a PDF buried in an inbox. They become a dated entry on the same timeline as the rest of the record.

Why it has to be the portfolio, not just the briefing

A briefing without a portfolio behind it is a summary. Summaries are easy to build and easy to commoditize. Anyone with an LLM and an EHR API can build a briefing in a week.

A portfolio is a different artifact. It requires every connector the clinic uses to land, dated, deduplicated, and reconciled. It requires the fax inbox and the OCR pass and the date-fixing for Fullscript order dates and the timezone-fixing for Oura. It requires the patient identity to be stable across six systems with six different MRN schemes. That’s not a week of work. That’s the work.

Build the portfolio, and the briefing falls out of it. Skip the portfolio, and the briefing is a parlor trick.

What we ship

The briefing is what we put on the demo. It’s what a physician opens 90 seconds before the patient walks in, and it’s the moment that makes the case for the seat. The portfolio is what we actually built, and what makes the briefing real on the second visit, and the tenth, and the hundredth.

If you run an FM or rural clinic and you’ve ever opened the room not quite knowing what the last six months of this patient’s chart said, the portfolio is what we built for that. The briefing is the demo. The portfolio is the product.