HNHacker News
TopNewBestAskShowJobs

seehafer

1,143 karma · joined November 22, 2008

Software engineer turned medical device product manager turned SaaS CEO. Now exited and trying to be a voice for little tech at the FDA.
submissionscomments
seehafer··on DHS Proposes to Remove the International Entrepreneur Rule
There’s probably a viable business model now in helping US-based startups set up offices in Canada.
seehafer··on Seattle now decade’s fastest-growing big city in the U.S.
Texas would beg to differ with this assessment
seehafer··on A woman’s brain stimulator stopped working after her home struck by lightning
Instructions for Use indicate that in such an event you should immediately go see your physician.

A pacer is different than a brain stimulator. And pacers generally try to restart themselves if they can. Brain stimulators, because of the sensitivity of where you’re stimulating (a few cm, literally, away from regions that control breathing, heart rate, etc), the least risky option is a shut down.

seehafer··on A woman’s brain stimulator stopped working after her home struck by lightning
I used to design these devices (deep brain stimulators).

As part of the design process for every FDA-regulated product, you are obliged to perform a risk analysis to judge the severity and probability of risks to a patient using your product.

One of the risks was “Exposed to a strong current that could destroy/disable the onboard electronics”. The design mitigation was a fail safe shutdown.

seehafer··on To solve affordability crisis, Bay Area housing stock must grow 50% in 20 years
I think Tokyo is a good counter to the nonsense perpetuated in SF that the law of supply and demand doesn’t somehow apply to housing.

Homelessness however is more a multi-headed beast, and Japanese culture likely plays as much into a lack of people on the street as does housing availability.

seehafer··on The Tale of Theranos and the Mysterious Fire Alarm
Theranos tried to get their tech bought by the military and used now SecDef, then Gen. Jim Mattis, then a Theranos board member, to shepherd the product through the military procurement process.

They were stopped because DoD regulations state that medical products used by the military have to be cleared/approved by FDA (or in the process of a clinical trial). And Theranos was not on FDA’s “good” list at the time, having only sent one product through the clearance process out of the dozens they were selling.

seehafer··on Many medical devices on the market have undergone no clinical testing
They would have to do biocomp testing. Every 510(k) requires an analysis of device compliance to ISO 10993, regardless of whether or not its predicate is preamendment. The premise of the article is alarmist and acts as if FDA requires no testing whatsoever, when the list of standards is as long as my arm: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpcd/cla...

Now one could argue that biocomp testing, which is done on animal models for the most part, is insufficient for any implanted device. That's a fair argument and one worth having. But the linked op-ed is a polemic.

seehafer··on Y Combinator is accepting applications for S18
If you’ve been rejected before don’t be discouraged, we were rejected twice before getting in last summer. Stick with it!
seehafer··on Launch HN: Enzyme (YC S17) – Automating FDA Compliance and Approval
Yes. This type of innovation — where you can demonstrate evolution on an existing product class — is very much in our wheelhouse.

And, yeah, FDA recently increased the range of condom sizes they would accept, thus enabling the kind of thing you’re proposing.

seehafer··on Launch HN: Enzyme (YC S17) – Automating FDA Compliance and Approval
> I totally agree with this. If you could create a software submission which efficiently communicated traceability, verification and validation info, and the other key metrics the FDA wanted to see, that would be pretty huge in terms of submission turnaround time. Best of luck guys.

Thanks! Sounds like you’re pretty well-versed in this, if there’s some way you can help you with a future product please reach out!

seehafer··on Launch HN: Enzyme (YC S17) – Automating FDA Compliance and Approval
> De novos, abbreviated 510(k)s, lesser used pathways.

De novo’s definitely, abbreviated 510(k)’s as well (that’s more of a requirements -> standard conformance challenge which a system like ours is particularly suited to demonstrate), but to your point they’re not that common.

> In my eyes it's really a question of how long FDA will be expecting traditional submissions while they roll out a model geared more around approval of companies vs products.

Yes but in a certain sense for many digital health companies this is already true, because they’re 510(k) exempt and FDA’s oversight occurs retroactively via QMS auditing.

The near term challenge for digital health, until precert comes online, is that there are a limited number of software experts at FDA. To the Agency’s credit, they have recognized this and are trying to remedy it. But the big problem we see with software submissions is that things get overlooked, and in such a way that speaks to an overworked reviewers. This of course leads to delays, which is why it’s very important for software submissions to be buttoned up. This is one the things we’re tryijg to make happen with our QMS, which generates more reports that clearly map from requirement->code->test.

seehafer··on Launch HN: Enzyme (YC S17) – Automating FDA Compliance and Approval
Took quite a bit of negotiating to get to a figure we could afford ;)
seehafer··on Launch HN: Enzyme (YC S17) – Automating FDA Compliance and Approval
There are a couple of things that are distinct about us contra an incorporation company:

1) Compliance is an ongoing activity regardless of how many new products you release. Think something like HIPAA, but more pervasive. So there’s an ongoing need that is ordinarily fulfilled by people - everything from a consultant for a garage startup to 1,000s of FTEs for a Fortune 500 company. This creates an ongoing, sticky customer relationship for anyone who can deliver significant value.

2) Our product and approach is valuable to more than just startups and enterprise life science companies routinely spend $10M+/yr on software systems.

seehafer··on Launch HN: Enzyme (YC S17) – Automating FDA Compliance and Approval
To your other questions:

Registration and listing: harder to automate because information systems on FDA’s side aren’t quite there yet. For GUDID, yes

CAPA, labeling, submissions: yes

Alternative submission pathways: what do you mean by this precisely?

Pre-cert: I think the premise is correct, in that the current quality system regulations aren’t well suited to modern SW development practice. But the devil will be in the details. Regardless it won’t affect us significantly because whatever gets released will have to have a phase-in for existing process that we’ll be able to adapt to.

seehafer··on Launch HN: Enzyme (YC S17) – Automating FDA Compliance and Approval
Hi all, happy to answer questions!
seehafer··on Litecoin has lost half its value since the creator sold all of his stake
Then again, going on CNBC and touting your coin as the next big thing and then selling your stake a few days later... there's a reason we've banned that kind of behavior in securities markets for decades.
seehafer··on Peak Pharma
Just like “peak oil” was undone by fracking this will be proven to be untrue by the wave of drug discovery startups that will decrease discovery time.

1) https://www.atomwise.com 2) https://www.recursionpharma.com

seehafer··on High Sierra macOS freezing and stops
Snow Leopard is arguably the best OS apple ever released
seehafer··on Where’s the Next Silicon Valley? VC Is Betting on the Midwest
Nevada side of Lake Tahoe? Easy. No space.

Reno? Easy. No ecosystem, aside from specialized things (Gigafactory etc).

I would LOVE to start an accelerator type thing just over the border of SLT, CA. But I would expect everyone would leave at the end of the it for a larger metro.

seehafer··on Ask HN: Huge enterprise customer wants to see our source code
Correct there's some (not necessarily daily or weekly) amount of sync frequency that you negotiate.
seehafer··on An $814M Mystery Near the Heart of the Biggest Bitcoin Exchange
You are underestimating the extent to which a finite $200M injection can make it seem like the price is skyrocketing for exogenous factors and therefore yield real money coming in. Classic pump and dump strategy.
seehafer··on Ask HN: Huge enterprise customer wants to see our source code
Seconding the “source code in escrow” solution as a mitigation about concerns of working with a startup. IMO enterprise customers happy to pay for it and it usually never gets asked about again
seehafer··on The Bear Case for Crypto, Part II: The Great Bank Run
I think that actually exacerbates the psychological dynamic that the author of this post is talking about. Not being able to get your money out will drive additional panic.
seehafer··on Ask HN: What tech were you convinced would take the world by storm but didn't?
> CRISPR technologies are like bullets to attack problems

This is a correct and useful analogy. Nonetheless, biologists are excited about CRISPR because, to continue your analogy, we were previously working with spears. And now we have bullets. Still don't know where to quite fire them yet. But it's a nontrivial advance.

seehafer··on All 50 startups from Y Combinator’s Summer 2017 Demo Day 1
Right, what I meant was, if someone wanted to see your video, they'd either need to ask you or be an investor invited by YC.
seehafer··on All 50 startups from Y Combinator’s Summer 2017 Demo Day 1
Yeah, you either need to be an investor or make friends with one of the founders and ask really nice
seehafer··on The Myth of Drug Expiration Dates
Unfortunately, ALT is not a solution, because it's not accepted by FDA, except as a temporary measure.

Any drug submission that supports shelf life via ALT data will be expected to also say "and we're doing real time testing to support this", and FDA will expect the follow up real-time data.

seehafer··on Babylon.js 3.0
Interesting, I get ~28fps on Safari w/o measurable increase in CPU utilization

On a brand new MBP, though. Maybe that has an effect.

seehafer··on L.A.'s crisis: High rents, low pay, and $2k doesn't buy much
LA fixed that by taking Sierra water https://en.wikipedia.org/wiki/Los_Angeles_Aqueduct

Vegas relies on the Colorado River, which is very overtaxed for what it supports now (Vegas + Phoenix)

seehafer··on L.A.'s crisis: High rents, low pay, and $2k doesn't buy much
Having tried the "Contact the City Council of a city I don't live in" thing in the past, I'd suggest a better use of your time, if you don't live in Brisbane or MV, is to try to find companies that do and persuade their leaders to take action.

Larry and Sergey could fix MV's intransigence very fast. We need to call on tech leadership to take a stand here. It's in their best interest.

+1 to all the suggestions about calling your city council, your assemblymember, your state senator, etc

← PreviousPage 4 of 10Next →