Ask HN: What is the biggest untapped opportunity for startups?

There are some very good reasons why startups flock to the bay area, including "lots of available talent" and "that's where the VCs are", but there are also problems with being in the bay area -- talent is considerably more expensive (due in part to the cost of housing) and visa issues (particularly under the current presidency) being the first two which come to mind.

If you can find some way to give non-San-Francisco startups the same advantages that San Francisco startups have -- better tools for remote workers, for example, so that companies can easily hire from anywhere rather than needing to be where the largest number of potential employees are found; or something to make VCs interested in investing in companies which aren't within a narrow radius of Sand Hill (since I've never dealt with VCs, I have no idea what such a solution would look like) -- then you'll create a huge amount of value for companies around the world and it should be easy to transfer some of that value into your pockets.

In my opinion, tools for effective remote working are there. Github, Slack style communication, collaborative project management tools, CI tools in cloud. Design tools could be more collaborative, but it ain't a showstopper. Video conferencing still sucks now and then, but 50 years after the Mother of All Demos, it starts to be usable enough :)

I feel that it is something else, culture issue, that still makes teams that are physically in the same location, to perform more effectively. Fixing work culture to support remote work better is likely the key instead of tech and tools.

And money will follow: when VCs believe that remote teams and remote networking are as effective, they will invest everywhere.

Corollary: It's probably time for a much wider adoption of bitcoin in tech. To transfer that value from halfway around the world to personal pockets capitalism style, we need crypto to become a primitive member of apps, markets and tech deals, and no longer a big deal.

- Spatiotemporal analytics usually in the context of IoT. Most people currently repurpose cartographic tools for this purpose but the impedance match is poor and the tools are seriously lacking elementary functionality. There is no magic technology here, just exceptional UX/UI and an understanding of the problem domain and tooling requirements.

- IoT database platforms, no one offers a credible solution for this currently. Everyone defines this in terms of what they can do, not in terms of what is required in practice. There are many VCs currently hunting for this product but the problem is one of fundamental tech; you can't solve it using open source backends.

- Also for IoT, ad hoc clusters of compute at the edge being able to cooperate for analytical applications. The future of large-scale data analytics is planetary scale federation for many applications. Significant tech gaps here.

- Remote sensing analytics. Drones and satellites are generating spectacular volumes of this data and no one can usefully analyze data of this type at scale. Today, companies wait weeks for a single analytic output on less than a terabyte of data.

- Population-scale behavioral analytics. Many startups claim to do this but none of them can actually work with relevant data at a scale that would deliver on it despite increasing availability of the necessary data.

- AI based on algorithmic induction tech i.e. not the usual DNN and ML tech everyone calls AI. This is way more interesting if you have a novel approach.

I am co-founder of tensorflight.com. We do computer vision analytics of drone imagery. Interesting that you mention it, as I thought it's a somewhat obscure market. When we talk to investors in the valley 75%+ have to be educated about why what we do is a viable business.

Please get in touch at kozikow@tensorflight.com if you have any ideas!

Any chance you can elaborate on what you're talking about here?

I know it's fashionable to hate on deep learning, but algorithm induction is literally what deep learning does.

Thanks in advance.

- An exabyte-scale storage engine. Nothing too exotic here technically and a few companies have built them, but the design needs to address continuous data corruption, continuous hardware failure, geo-federation, etc.

- A real-time database kernel that supports very high throughput for mixed workloads. A production kernel of this type doesn't currently exist though several people working in closed source databases understand the necessary computer science in principle; the academic literature is far behind the state-of-the-art. The ability to gracefully shift load and transients between servers under full load with many millions of writes per second is not trivial.

- Native discrete topology operators. Necessary for geospatial analytics, sensor coverages, etc. If you can do it natively in the database kernel, it makes the second requirement easier to achieve since you don't need secondary indexing generally.

Any solution even halfway toward the general solution would be viable. The value possible if you have such a system is hard to overestimate. Companies have paid half a million dollars for the output of a single analytic query on tens of trillions of IoT records; the differentiator was that it was possible to execute such a query at all.

It is extremely high-end and polymathic computer science, but serious valuable if you can make a credible dent in it. And unlike some advanced topics in computer science, there are no epic unsolved theoretical problems you have to solve, though some relevant computer science may be unpublished.

[0] https://a.kx.com/q/d/q.htm

[1] https://kx.com/solutions/utilities/

His company provides sensor systems for civil engineering projects. A single large bridge can have sensor packs every 50m or so - per beam. The amount of sensor data coming in for a single municipality or region is already staggering. Vibration and stress analytics are required on a daily basis.

The final requirement, one which you didn't mention, is that this setup should be fairly low maintenance. If you need a team of rocket scientists to operate and just keep it from falling over, the cost structure will be unsustainable.

A service that provided all this in a platform with sane APIs and good BI integration should be making tons of money.

I'm just speculating, of course. I was sort of entertaining your idea and going back and forth between "a sounds-good fantasy" and "No, that would be awesome, why does this not exist?"

Three big missing features off the top of my head:

- Insufficiently correct and high precision computational geometry, which compounds with the iterative/recursive nature of many complex sensor analytics. Many people don't notice unless they ground truth their analytic process; I learned this the hard way. For many industries, 1% cumulative computational error is a catastrophic bug for analytics and the reality can be much worse in many common systems.

- Lack of first class tessellation types and operators. Once your data scientists have them, they'll wonder how they lived without them. Such things are completely useless for cartography and therefore don't exist in those platforms.

- High-performance computational geometry. This is particularly noticeable if you work with sensor coverages (like drone data). Your typical cartographic system has serious difficulty joining a few terabytes of complex polygons, but these are tiny data sets for many remote sensing sources. It literally takes weeks or months to run these types of queries. You can optimize this to be much faster but there was no pressing market need in cartography and cartographic systems aren't designed for scale-out generally.

Crytpo Currency: There is room for more disruption here. I suspect a currency that is both trackable and backed by a pool of commodities/currencies could be quite popular. Traceable would make theft risk reduced as money could effectively be returned if it is stolen and being backed/hedged by currencies/commodities would help with confidence.

Cargo: I'm surprised we haven't seen electric cargo ships. Even combine solar with sail as winds are favorable. This combined with auto-navigation (at least between ports) seems more easily achievable than cars yet technology is further behind.

Dockable Phone to PC (physical or even better if wireless dock): Surprised no-one has done this well yet. I can image whoever does this with really take ownership of the OS space. I always felt this could be the best route for Microsoft to re-enter the mobile space with force.

There could be could add battery storage to be used during the journey. They could include wind turbines to add further power generation.

Also ships can go slower. In my limited knowledge of boats every extra knot takes significantly extra fuel. If we were bringing multiple power sources 'slow' cargo may be viable. And the has to be other ways to increase this function.

And there is always hybrid. Not all power has to be renewable sources. A solution could start with supplementing diesel driven thrust with some lower cost green power. Given diesel is 70% of shipping operating cost that seems a potential option.

No one really wants to build cargo ships right now, shipping companies are barely staying afloat right now because there is a glut of cargo capacity.

This technology would need to be cheap enough that it would make sense to retrofit a ship with it... which would probably not be the case for a very long time.

I do think it's a very thought provoking idea however.

I feel its like Palm vs. iPhone. Sure Palm was earlier to market with a reasonable solution, but someone needed to nail it.

While it's easy to say IoT, cryptocurrency, or whatever the latest buzzwords happen to be—there's ideas that have been floating around for years which are still viable, it's just that they're hard and require exceptional execution. In that sense, they are almost timeless until implemented correctly.

For example, another comment suggested marketplace/content discovery. That's been an unsolved problem for almost a decade now. Ads are another great example: they've been dishing out human misery for about the same length of time. People hate them, so they use ad blockers, and everyone loses. These aren't new problems or opportunities.

Also, I can't help but be struck by the similarity of "mainframes -> PC -> cloud".

I'm a little sceptical of first party content sensitive untargeted advertisement, but we'll see. I also think the people who would block ads would block these too.

See responses to this comment https://news.ycombinator.com/item?id=13576611 where people say they will block ads anyway. It's not the tracking. It's not the bandwidth. It's just because it's easy to do.

Old Navy doesn't want to deal individually with a site that doesn't have 10,000,000 visits/month. So we end up needing an intermediary at the same scale as the big advertisers.

Not saying this can't work at some level, just that there are some big economic forces fighting against scaling a product like this.

I would think this is the main contributor to why we have shitty ads. Everywhere.

Seems like there's room for a move something like what DigitalOcean did in the VPS space.

When EIG took over Bluehost the first thing they did was outsourced ALL tickets to India (Hari the CEO's parent's company) - Then he moved chats to India. Then he closed down tickets completely. Chats may eventually be on the chopping block. Then he moved sales to Tempe and promised everyone in Orem that their job wasn't going anywhere. Two months later he announced -- hey sorry I lied, we're all moving to Tempe. You can come w/ us if you move yourself, but here have 1 month pay for severance. I'm pretty sure I just heard that if you work for EIG and don't live in Houston, or Tempe you can count on your Brand closing and moving to one of those two places - they really want to consolidate everything in a bad way.

i envision a company -- where Each employee has a book of business, each employee is also an affiliate and can refer business 24/7 and get commissions for life off anyone they refer. Each employee acts more like an acct rep than actual CSR agent grinding out phone calls left and right. Customer's have a dedicated person they can reach out to with any concerns. There would be very generous ESOP plan, it would be setup a bit like maybe Winco Foods, etc... and other bonuses. Exec pay would be capped at 65x avg salaries. Surpluses left over go into bonus pools and activities for employees, etc..

You'd need to combine an ISP and MSP with an add-on like analytics and integrations.

Trouble is, then you're competing with AWS and GCE.

You have what amounts to bare metal through AWS, available by the hour. You can have that today, plus the ability to scale across geographies at the click of a button. Few colos can offer that.

I have 20TB of data that I would like to access about 1MB statistically per year (!!) only I don't know which of that 20TB its located. Sure there is AWS and then Glaciar but with my limited knowledge with AWS you need to spin a computer and then set the rules how those files are accessible and it already goes into tens or hundreds of bucks per month. Amazon Glaciar is too expensive to pull data out.

Perhaps I didn't do enough research. If you know company who allows me to cheaply store XX TB in cloud and charge me only per access to it, let me know please.

Thanks for the recommendation though.

Disclaimer: I don't work for MapD.

About your comment, imagine that the app can know my location, and pair me with a local expert, if the expert has to go to the other person home to fix their computer the price increases.

- Twitter w/out fake accounts.

- a marketplace that uses Facebook as a vehicle for engagement/promotion but which operates independently.

- Secure SMS for 2FA tokens

- Android w/out Play Services

- Schema-based email

- Stripe for the rest of the World

I might postulate that programmers have more of an aversion to chatbots than the general public. Presumably because we tried one of the famous general purpose ones at some point - then were disappointed when we failed to make it understand why humans cry and why it is something they could never do.