The market size for international trade logistics is staggering...you have at least a hundred global logistics providers with market caps over $10B, and more than a handful of which are in the $>100B territory. And they have trillions in revenue combined. And they are in a capital intensive industry on low margins, and they dump tens of billions every year buying companies that can give them slightly better competitive advantages. And they ALL suck at regulatory compliance.
But here is the bigger secret: The direct monetary cost of compliance isn't the big pain. What these companies pay for is delay reduction. Customs delays can drastically affect end-to-end SLAs, have a huge impact on customer loyalty (If your shitty compliance led to my product being held in port for a month, I will never use you again), and they have massive impacts on capital utilization due to how difficult it is to plan in the face of seemingly random customs delays. For an industry that is judged on Wall Street almost purely based on capital utilization (return on capital), you can bet your ass that every single one of those companies will be throwing as much money at you as they can.
- Electronic source data is often unavailable
- Each country has unique and often conflicting processes
- Every company has unique and often conflicting processes
- Every port has unique and often conflicting processes
- Deployments of truly automated solutions are easily as complex as some ERP implementations
- There are many cheap point solutions to compete with
- Compliance heads are often more comfortable increasing headcount rather than buying software
- Experienced compliance professionals are crucial to development
- Experienced compliance professionals who understand and care about software are few and far between
- Logistics and compliance professionals do not play nice
- Very driven by the regulatory environment. Nobody wants to pay for compliance until they get fined.
- You'll also be expected to interface with multiple government agencies, none of whom are overly concerned with your business success.
However the format they've created is already outdated (and it's taken three years to get to initial draft). Like most large parcel companies, they are catering for slow, technologically lagging suppliers and customers.
It'll take a bleeding-edge integration partner like uShip, EasyPost or GoShippo to create a standardized format. (EDIT: Or the likes of Google who managed a similar outcome for addressing data with the Google Maps API.)
This is really a field that large government bodies need to address seeing as most governments still have a large stake in parcel shipping.
EDI formats still dominate though, especially in logistics and on the government side.
Just a thought.
You'd need dedicated people to go through stuff at individual major ports, speak the language, know all the people, etc. As you rightly point out, trust is paramount, and I think trust absolutely requires people on the ground.
For everybody else, FedEx/DHL is hella expensive if you want to ship a couple conexes worth of stuff.
The way freight companies are set up, you can't just pull up with a container and say "What's the rate to Albuquerque?" you have to have a whole plan that takes into account transfers, mode shifts, delays, temporary storage, customs, who can sign for it, paying every single entity involved in touching the freight, and a bunch of stuff I don't even know about.
How does it get from the warehouse to the port? Going by ocean or air? What taxes, duties, etc. are due at customs? How does it get from the receiving port to the final destination? Who arranges each step? Who pays for these things?
Things happen like shipping containers showing up on the docks without a proper cargo manifest or other documentation errors, which of course leads to the shipment sitting in customs until it gets sorted out. This is because the humans creating those forms often have poor tools, and of course there's the usual factors like staff who are too busy, incompetent, or lazy (in an ideal world they should not be doing that job, but this is reality we're dealing with).
The obvious answer would be a shipping management system integrated with the company's supply chain software (typically, their ERP), but usually these problems are coming from OEMs, subsidiary companies, and occasionally 3PLs that have aren't part of the company's systems. Any integration with those systems is usually rudimentary at best, and it's not unusual for the "integration" to be human-based.
1. Acquiring Industry knowledge
2. Getting a sale
3. Developing the software
We manufacture ductwork products for heavy industry (sugar mills, power generation). Each company has its own supplier registration process for procurement, usually it involves entering our company's basic financial info, customer and project references list, etc. into some kind of custom Java web application.
There should be just one place that asks us to enter our DUNS number, our tax id, upload our W8BEN form, our ISO certificate, our company contacts, our references, our company financials, etc. and then mass-submit to all supplier registration forms within the industries we want to target.
This is a really simple business problem with a clear technical solution. Honestly you could just solve it by:
- WuFoo form to collect the company's information
- Excel spreadsheet to list the vendor registration portal URLs for each company, with different industries in different tabs
- Human labor to go through and find the data from WuFoo and plug it into each supplier registration webapp.
You could make money on a pay-per-registration model. All that would be required would be a simple portal for users to see the progress of each supplier registration application. It's basically CRUD.
Maybe some ambitious company would try to innovate by building the One supplier portal to rule them all, but for now, this is an opportunity for a quick win & a business that could be run for cash.
At the place I worked we had 3 different supplier databases, each one having been introduced to replace the last, which was never fully phased out.
Imagine there's another company, let's call it Texas State Power Construction (TSPC), that needs a new widget to replace the old one in the plant. The engineer at the TSPC plant sends a request to the procurement manager asking for a new widget. The procurement manager then searches his or her (more often her) Supplier database for registered companies that make widgets. Procurement manager doesn't contact Acme because Acme never registered with TSPC. Acme loses out on this business opportunity.
Suppliers get in the db by registering on the enterprise-y, most likely Java, "Supplier Registration" portal linked to on the website.
Registration is a pain in the neck mostly because it's boring and nobody wants to do it. I would love to do it for every power company in the USA, but it's boring. I guess I could try to hire an employee to do it for me at $15 an hour but I would rather just pay for results.
(My company is an "Acme". We make a "widget". We would LOVE to be in the database of every "TSPC" and get more sales inquiries.
I spend $500/month on adwords and get maybe 2 leads, and they're of dubious quality. $250/lead sound too high? The last trade show I went to cost $1600 in expenses alone and I generated only two good leads. That's $800/lead.)
@mdolon I sent an email with some of my thoughts mapped out in greater detail. Thanks for asking.
Ever log every single sensor output (as well as 200 more) every 50ms of a vechile over an 18 hour road test while recording audio and thermal video from 2 dozen different points? How about on fleet of 40 vehicles daily, for 2 months, in the middle of Alberta or Death Valley.
Have a good way of querying, analyzing, processing, and securing all this time series data in a way that can handle literally getting >100TB per hour? And can keep up with the expected geometric growth? (I've got a decent solution for this actually currently needs to be vetted). (Also security has to provided on a per-channel basis, not per-test, T1/T2 companies need access to their test data, but not global data).
Contact me. I'm a member of the ASAM standards committee that recently met to discuss how basically every auto-producer and tier 1 has NO CLUE how do implement this. And easily 2 dozen companies are just waiting to throw money at this problem.
Currently one doesn't exist, and the solutions that do exist manage paths to raw data blobs, not actual records/data points.
Talk to companies like Honeywell (I believe they do something similar to what you are trying to do for aviation industry), ABB, Foxboro, etc. These guys have been doing this since 80s.
Feel free to get in touch.
My email is in my profile if you'd like to chat. :) Hope to hear from you!
Is this data logging the changes or the states of your sensors? If it's the states, then I am guessing most of this is highly compressible. If it is actually 100TB of changes logged, then that's a pretty difficult problem.
Here is the related HN thread:
This method is already used in data logging compression (slightly) where one stores channel delta's/time stamps. Reconstructing the value ad-hoc when necessary. This is a good way to compress non-violalitle datasets.
I've actually watched the talk already. And while it seems to apply the problem is it doesn't. Every data point is important, because the real problem is comparing different tests, with time between tests to attempt to get an idea of how hardware ages. Or to test componenet swaping, where a known test is performed on several different items and in post processing the results are compared. To use the suggest method your storage solution requires knowledge of whats being stored.
The goal is to unify these storage solutions, and present a unified front end for querying/report generation.
Could you expand a little more on what sort of features you would like to see in a solution? I have some relevant experience and I could see myself taking a stab at this problem
If you pulled that together and offered it in a usable format.. Wow.
You have no contact listed...
I really miss the challenges of real-time signal and data processing. I really want back into it.
Think of a soccer stadium, with fans taking "video" of the game. All the feeds would be gathered, synchronized, quality adjusted and put online for anyone to view, from any angle.
Even the buzzword-du-jour is getting involved.
I know the technical side/requirements, and the people. My business skills and design are lacking. Also a whole front end for interfacing/formating/report generation needs to be created.
Right now the process is pretty much that you (or someone you pay) have to do a ton of reading and become an expert in import/export and RF law for any given country which you wish to sell or distribute a physical product in. Then you have to get written quotes from a few vendors who you find that actually do said testing and generally the test houses are just test houses, they don't necessarily know the laws of each country they just know how to run the tests and issue you a report, which can take weeks, at best (or you just go to the local place and it costs what it costs). But you have to know what tests to ask for.
There should just be a simple one-stop online form that you fill in your product's information and get an instant quote back with each country in the world and what's required for your product along with instant quotes (price and lead time) for each of those certifications. Then, you can check box all the certifications you want right now (likely with discounts if you pick more than one, so a live-updating pricing matrix will be useful here, and a live world map highlighting countries which you can sell in) and pay via credit card for the certifications themselves. You get in return a mailing label to ship your prototype device to the testing company along with instructions on how the testing company uses your device so they can test it.
The actual testing can be contracted out to 3rd parties. This could just be a broker, but it would have to be a broker that knows the rules of the world and can express those rules in a straight forward way along with instant pricing. The prices don't need to be the best prices, they just need to be straight forward and competitive.
However, my experience with the test house has been different: (1) they actually do provide the knowledge and experience with regards to your specific product and needs, (2) it takes a bunch of conversation and bi-directional education to get the manufacturer and the test house on the same page, and (3) it's still fairly unique to each product. For example, we had to build custom firmware and custom hardware just for the testing process. (Remember, this isn't dealing with an efficient marketplace; this is dealing with multi-country government regulations that are designed by big companies to erect barriers to entry for smaller competitors.) And the volumes and willingness-to-pay from small hardware startups are low, and the test houses make most of their money from bigger companies that do more product variations, more iterations, etc., which requires relationship building, rather than a one-stop online form.
So, while I'd really like your version to exist as a customer, unfortunately I don't believe it's a match for what most of the hardware testing market looks like.
Obviously the broker I describe can't know everything about your product and market segment and customers and said laws that would apply, but 80% of the time, there's nothing unique to a product from a regulatory point of view, someone already has a product which falls under very similar guidelines. For example, if you don't have a radio transmitter, the types of tests you have to do for FCC are pretty straight forward. And if you do have a radio transmitter in an ISM band, again the tests are pretty cookie cutter. And if you aren't AC mains connected and don't have any voltages above X VDC then UL low voltage testing is easy. Lots of products would fall into these categories.
It's when you venture out into niche market segments with radio transmitters that you really need to know what you're doing and not rely on a test house or broker (besides, your company is liable no matter how much info the test house or broker can provide you with).
Yeah, if you're making a cell phone or fancy internet of things radio device, this isn't a necessarily a service for you. But if you're taking an arduino-like prototype and turning it into a real product or making a new USB charger or a keyboard or computer input device or ..., then it could fit quite well I think.
The answer was that our doctors are disconnected from the costs of treatment because they can't be expected to know every possible charge for every different insurance plan (what the patient would have to pay for procedures, medicine, support staff [anesthesiologists who are out of network, for example], devices [pacemakers, wrist splints, artificial hips, etc.], lab fees, etc.). So why not something that can bring that information to them?
This would be a benefit not just to consumers but to insurance companies as well, and that's where the money would be.
Feel free to cut me in on the action.
The difference is in insurance:
1. You pay directly for your pet's care. Your vet gets the money. No in-between to obscurify the money.
2. You pay insurance for your own care, and your doctor bills your insurance. Your doctor's EMR might know the insurance cost, but that cost can still change. And there are lots of middlemen in billing processing in healthcare, so it's not an accurate representation of reimbursement anyway.
There are quite a few startups trying to remedy this in various ways. Most of them focus on data transparency; they try to aggregate and publicly list all the direct pay costs for different doctors. The problem with that remains the same: most people don't have the precedent in their head of paying for healthcare, so they don't want to pay direct.
Not impossible to fix. Someone certainly will. Probably orthogonally, like how Slack is unbundling email. But the tech barrier is small compared to the social barrier.
Would be pretty sweet if someone was able to map out all insurance costs though, even to 90% consistency. Hooo boy that would be sweet.
I'll stick around the comments too but email me email@example.com anytime if I can help with this madness. Been through it way too many times myself.
I was diagnosed with a melanoma a couple of years ago.
Luckily early stage but had to deal with several hospitals and dermatologists who just doesn't talk with each other, add to that the complexity of insurance-agencies and the fact that I have more than a thousand of those little bastards and you begin to see the potential for mistakes and the stress that might incur on a lot of people.
Whoever could solve this would be someone I would pay money.
This is probably a difficult problem to solve, but one thing that might be possible is just a way to look up the bill and translate the insurance "codes" that are meaningless to normal people and require looking up in various medical or insurance sites. Like code 123456 means you had your blood drawn. Or code 67890 means they ran some test, etc. It takes forever to look this stuff up.
You're also talking about going to see the Dr for a cold or something. When you've had a major surgery and months of follow-up procedures with pages and pages of line items on your bill, hundreds of codes, disputes with things you think you have already paid, bills that don't match up with what insurance says, etc... then you'll understand what I'm talking about.
On the other side, its very hard to find good products online since the search results are full of SEO optimized junk.
If only there was a way to combine these two together. e.g Users could enter what they are looking for in an app and they only get very targeted advertising from business big or small inside that app. Matchmaking algorithms help make the initial connection. Think of a combination of Magic and Tinder.
I think that's the actual business model for these companies. there's a reason I don't accept those pitches.
That made me laugh.
> "SEO optimized junk"
This did too.
>"only get very targeted advertising from business big or small"
Wait, did I misunderstand your first points?
Seriously, is this post a joke or were you serious? I cannot tell.
You can't have it both ways. I would say that a good portion of the ads I see are surprisingly relevant, but I might be biased because as a marketer, the people targeting me are the best marketers out there.
A place where I, or anyone trying to contact me, can send snail mail to. These then get scanned and forwarded to mail mailbox and/or dropbox and/or stored online with an OCR PDF.
This would allow me to
a) send my bills, invoices there
b) able to move house more easily without having to keep track of address changes.
This has caused during that snailmail delivery will be only 1-2 times a week, when it now is every workingday.
Facebook doesn't seem to work too well, and adwords felt like a black hole. And besides these methods seem to require experienced consultants to run an effective campaign.
I have seen dozens of business start off this way and than eventually become enterprise only platforms or worse, programmatic advertising providers marking up CPMs by ridiculous ripoff margins.
(see: boost media, dispop, etc...)
That's why we built Hits Analytics (http://hitsanalytics.com) specifically to tackle this problem. All you need to do is set the budget and pick a segment of product(s) you wish to promote, and we will automatically create and track Facebook ad campaigns for them (we do the "heavylifting" of optimizing the campaign behind the scenes based on the conversions we are tracking)
I know Amazon tried to solve it but failed, I guess a few of those deal sites too are trying. Right now the only true contender i guess is something like graigslist?
The market leaders right now are God awful. Market leaders include Adeptive (Resware software), ISGN (Gators software), and Ramquest (Closing Market software).
Their platforms usually include a desktop app (not really necessary), in most corporations accessed via Citrix, or installed locally (host of issues with versioning), and an accompanying database and web application. The web app usually offers a very limited view into orders, and some limited ability to place orders and is customer facing. The desktop app is usually internal to the title company.
The amount of money spent supporting these applications is staggering. In one company I know, it's ~$3mil a year in application support alone. That cost alone would make a strong case for switching to something more intuitive. The worst part, however, is that they're also extremely glitchy. When someone misses a closing because of this software, it's a huge deal. That's somebody in a hotel somewhere, and a very unhappy bank.
Some of them suffer from insanely limited UI's, application design that is prone to database blocks, and various bugs such as accounting data updates not being wrapped in transactions and failing to update 2 tables when needed (so you get 'half' of a payment and your ledgers are out of sync).
In practice, here's all these boil down to: It's a task list (provides user with a list of tasks to complete) that has some attached functionality to help you do the task, and is customizable. The reason most are desktop apps is because the most used functionality is document preparation, which seems to be universally done in Microsoft word using templates by all these companies. They're also horribly dated.
If this could be a web app, work well, and port data over from the current market leaders into your schema, you would not have to work hard to convince title companies to switch.
I'm setting up a brand new postgres solution to replace an existing highly-tuned MSSQL solution, and the pain points are all down to differing performance profiles. I have an SeoKeywordsProfile query that takes 3.5 seconds in MSSQL, and >12 hrs on postgres. Clearly there's an index that would solve my problem, but the query is 3 pages of SQL with no whitespace!
There's https://github.com/cohenjo/pg_idx_advisor, but it only works on one specific release of postgres and not on any other.
I'd like something that could just monitor my database for a fixed period of time and tell me what indexes aren't being used, what indexes ought to be used, and make any other suggestions that might help (partition this table, vacuum that table, etc).
Set up pgBadger https://github.com/dalibo/pgbadger and see https://wiki.postgresql.org/wiki/Index_Maintenance for more information about your indexes.
I tried that.... http://explain.depesz.com/s/22H5
The problem is, I'm unsure how to read this. I can explain, but I can't explain analyze, since it's too slow. Can you make heads or tails of it?
Lines 30-32 indicate that you are pulling a lot of data from buyclicks and transactions. This makes up a huge portion of the cost of this query. Lines 33-40 are very similar. Looks like you are probably pulling more data than you need, and you'll see a lot of improvement if you can reduce these two spots. Fixing those two parts will make all the nested loops and hash joins that use this data a lot faster.
Looks like there are several sequential scans (Seq Scan) that you could likely improve with a quick index on the relevant columns. You seem to be filtering for similar things on ad_codes in a main query and a subquery, which probably could be refactored out. These are small optimizations compared to the first.
My email is in my profile if you want to talk about this more.
Hash Join (cost=78,396.37..2,150,742.13 rows=24,442,180 width=40) (actual rows= loops=)
Hash Cond: (buyclicks.offer_id = offers.id)
if you can filter down number of rows in these cases or maybe only return columns present in the index and/or add an index where required you will speed up. Watch out it doesn't affect insert/update too much though if you add an index. You have created fk indexes right?
Yes, we have fk indexes.
I don't see how we can filter down the number of rows, since the purpose of the outer query is to get a count of how many buyclicks there were for each offer. But I'll see if I can create a covering index that reduces the join cost.
More than that ... I want somebody to make an open and hackable ALM platform that I can build my own tools on top of.
The product should be license-free (ideally but not necessarily FLOSS), so I can put it on my CI servers and VMs without worrying about license files and license administration. (Flat-fee site licenses would be OK, but anything that stops me from freely cloning VM build-machine images is a big no-no).
I want diff-able plain-text-backed requirements management, defect tracking, and project planning, with everything (or its hash) co-versioned alongside the source, so I can use my VCS to provide a coherent, branch-able, change-oriented view of project state and project history.
I want the text-based backing files to be a simple, accessible format like JSON or YAML, so I can write quick hack-tastic Python scripts to build my own development analytics and reporting tools.
I want all this because I am fed up with having to do it myself ... and because I would rather spend my time writing tools to analyze classifier performance and tune algorithm parameters without having to do so much book-keeping with difficult-to-automate requirements management and configuration management tools.
For example, I would love to talk to a series of very experienced (read older people) people working in various industries what they consider the biggest problems they have to deal with. People who are on the verge of retiring.
This is why I asked the question this way, so far some really good problems to tackle.
Finding trustworthy travel destinations for families that don't like all inclusive but prefer smaller boutique hotels that are family friendly.
Despite all the online travel sites that are out there its still often hit and miss how great the places we go are.
sure theres plenty of hotel or airfare or rental or 'town of xyz events calendar' sites but no one solving the entire problem.
I've yet to find an end to end trip planning site, especially one that is family oriented.
My guess is that the solution is something simple, something that isn't algorithmic but rather a platform for travel agents who can listen to and understand the individual needs.
I would say the Atlantis, Bahamas resort does a good all-inclusive. It has activities for kids and adults. The resort I stayed in Malta was clearly for young adults. The kid and I were practically chased out the pool once the sun set.
We are looking for those that are but without the tediousness of the all inclusive crowd.
Any suggestions on family friendly places you've found?
I have stayed on many expensive hotels but that is not what we are looking for. We are looking for hotels that are family friendly so other families go there but without it being a family hotel.
We will pay whatever it cost but it's not so easy to find that the balance.
Lots of small businesses can unlock significant value with better software, and could be convinced to pay for it. But the upfront cost of acquiring those customers is too high to make it worthwhile.
If you could offer a highly-curated marketplace, you could spend the customer acquisition cost once and sell to them over and over again, so they're buying all their IT through you, trusting you to vet the products and get over their risk-aversion. You could sell bundled solutions that are guaranteed to interoperate out of the box, lowering per-transaction integration costs.
SMBs are price sensitive, so charging what it costs you to deliver this will decrease adoption and subsidizing this cost can take too long to recoup.
A good example of this is Office 365. Let's say your profit margins are $2/user/month and you spend 3 hours onboarding a new 10 seat customer at a cost to you of $1/minute on the phone. It will take you 9 months to recoup your onboarding cost ($180 in cost / $20/profit/month).
Self service adoption can be possible depending on the product if there is no need for integration or legacy data migration, however products that require these activities see a high rate of drop off in the self-service channel because SMBs often times aren't sophisticated enough to solve these challenges themselves, don't use the service and churn out after a couple of months.
Additionally, if the customer has a poor onboarding experience with the first product they've purchased you lose their trust and it makes it very hard to cross sell and upsell other services.
I'm curious how curation would work.
If something is going to be plugged into the wall, it will be badly fitted, upside down, unplugged by the night cleaners for her Hoover and when you are on the phone to the engineer he won't know his arse from his elbow.
Get good young people in a dozen countries - give them decent wages, half decent travel options and a training and upgrade path.
The problem is that the pie is so small and there are so few of them, it would be tough to make any money.
But musicians hate selling performances, and venues hate buying and managing talent.
Ah, so that tends to explain the parasitic phenomenon known as Ticketmaster.
If they'd all just subscribe to a service, then they wouldn't have to dicker with a bunch of musicians and have their tasting room GMs deal with booking folks.
The same thing could apply to all the little bars and restaurants here; they have a reoccuring need to hire talent, but they usually have some staff member who doesn't want to be doing it (and isn't getting paid much) buying talent, so it's a pain in the butt for musicians and the bars get inconsistent products.
But at that end of the market, the budgets are so small that it'd be tough to take much out to work as a service.
But problem I am facing is my users are from all over the world and every country has different taxation law. And there is no proper service to pay my users except paypal (whose credibility is getting worse day by day).
Also, find a way to generate passive income for FOSS developers and crowdsource FOSS projects and turn them into decent products (proper doc, saas api, maintenance/support etc.) .
Not affiliated with them, but was a happy customer when I worked at a real estate company recently.
Large catalog where every screw is of different shape and requires a different tool to be put into a different wall in a different town so to speak.
Many companies (including Google, etc.) are making healthy progress here, but nobody has managed to really tell me the incremental value of my display efforts, the value of a view-through, whether bidding on that brand term really made a difference, etc.
Right now the tools out there can give you different lenses to view data through (static/custom attribution models), or take a stab at finding a data-driven, dynamic attribution model.
However the outputs all seem to be pretty underwhelming and do not typically spit out actionable insights like "Increasing spend on X, Y and Z placements should give an incremental revenue lift of $$$$."
Companies have tried, and for the most part failed. If anyone thinks they have the solution I've been waiting for, please don't hesitate to respond. But as a starter, your offering should do more than the free version of Google Analytics.
What? A tool that helps discover inefficiencies in Hive/Presto/Dremel query/pipeline/scripts.
It is so much easier to just add new machines to your cluster, than to optimize your code and fix inefficiencies. But the latter option typically results in millions of $$$s in savings.
I have owned a single family house in the states for five years. Just put a new roof on. Big expense. Used to live in a condo where a reserve study had been done, and every month a portion of the dues were allocated to large upcoming expenses. They recently replaced the roof of a 100 unit building and paid for it out of reserves (no assessment).
Would dearly love to have someone come in and do a reserve study of my house, give me a figure for how much I should be saving every month. Possible add ons include property management services (lawn care, etc).
Currently most title/escrow companies are local brick and mortar shops, largely due to slight regional differences in how the process works and who pays what (these differences are mostly historical artifacts that we don't want to mess with) .
At Open Listings  we're building the slickest real estate offer/transaction platform in the industry but it involves a fair amount of offline coordination with other agents, inspectors, banks, and title/escrow companies. Essentially, we're building an API on top of a lot of human interactions and an interface for our clients and agents to drive that API. We'd love to be able to plug into a preferred escrow service to do the title transfer and replace one slow, expensive interaction with an API call.
I noticed that OpenListings is looking to try to disrupt on the buy side of real estate transactions. Curious why you wouldn't try to do something on the sell side where I think there is massive room to change things and it's pretty darn formulaic.
All this only applies to the US market,not sure about international RE. Source: I worked for a brokerage for years.
Regarding buyers: we heard a lot of frustration from savvy buyers... people either didn't see the point of their Redfin agent or wanted a simple way to go direct with the seller. The idea of making it less expensive to become a homeowner also really resonated with us.
Foldable / Transportable Bicycles.
1. Build a bitcoin based, secure methodology for money transfer between individual/institutional NGO donors and end-users (cutting out a lot of NGO middle men and/or Western Union style folks)
2. Build a simple, deployable, open source system for low income countries where SMS is still used to find information on stuff like farmers wanting to know prices at new markets.
3. Micro-services to help young NGOs get setup.
-Find individual/institutional donors which match the NGOs theme
-Help with regulatory process
-Help with introductions to donors
-Admin tasks - website, brochures etc
4. Build a secure, Open Source platform for low-income country government management systems. (With modules, for example: Need a simple health management system? - here ya go...need a simple method of managing education results? - here ya...need a simple method of managing voter registration? - here ya go etc etc)
5. Build a system, ranking etc for measuring digital security risk (on a close to real time basis) on a country-by-country basis. The sort of thing already used on a physical basis for stuff like kidnap and terrorism...So I could travel to XYZ country and see that last week there ABC number of censorship incidents, ABC number of reporting fishing attempts using 123 exploits etc. This is only the first stage, the REALLY hard part is then linking it to tools, behaviour change and processes to mitigate these risks...(We are trying to hack together something a little in a very very simple version in our app, Umbrella but dam is it hard.)...
6. Make PGP so simple my grandma can do it (HERE, TAKE MY MONEY!!! :)
For instance do my grandma really care about cryptography?
By turning it around and ask for the problems I am hoping for a more open ended list where the solutions aren't given already.
As you alluded to, it all comes down to the difficulty of selling to them. No matter the size of the customer, they all seem to take two years to make up their minds to buy. That is pretty harsh on a scrappy startup. To grow quickly, you need to pour a very significant pile of money into an enterprise salesforce two years before you actually want to start making money.
That also leads inexorably to the conclusion that small customers aren't worth the hassle, so you need to target the behemoths who can spend enough money to justify their acquisition cost. And the behemoths are so scarred by years of disastrous IT projects that they are deeply risk-averse.
I also think the industry remains in denial about the strategic value of software, so you don't actually get paid what you're worth. They all know they need software, but they see it more as a necessary chore than an opportunity to revolutionize their business.
My hope is that a real technology company directly enters the healthcare market and starts kicking butt by building their healthcare services around great software from the ground up. That would shake the incumbents out of their denial.
I don't think their insurers would like that. When medical records start getting lost/exposed because of small outages or vendor changes, the Feds will come down very hard.
the healthcare system i work for has 650+ applications we support, I was hired as a developer yet we do very little development here. Hospitals like the idea of a vendor they can call and make fix anything, because its very costly to have SMEs for every single piece of software we use. many of them requires us to send our staff our for training yearly.
example: one vendor sells us diagnostic imaging software, but require us to buy a "certified" video card if we want their support... its just an old 90's nvidia card they rebrand and charge us $1500 each for.
Epic and Cerner are probably the largest EHRs. Epic revenues were 1.56B in 2013. While Cerner did 2.66B in 2013.
also, a lot of healthcare spending is government subsidized, which makes it even more lucrative.
http://www.healthcare-informatics.com/hci100/healthcare-info... this illustrates the money moving around in healthcare
- Setting up a RAID0 dev ubuntu box and transferring my data to it
- Setting up a VPN to EC2 for both our office (using our router that does VPN) and our personal machines (for when we're away from the office)
- Diagnosing office internet issues
Email in profile.
Right now, our transcribers manually do the typing. If we have a ASR based system, which produces even a 80% accurate transcript of the audio, we can improve our efficiency manifold.
We have a lot of data, audio files and their corresponding transcript. That data can be used to train a DNN based system. A software solution or framework which can do that would be ideal for us so that we can keep feeding in the data and continuously improve our model.
At one time I invested in a medical transcription business that experimented with voice-to-text technologies but couldn't generate any additional efficiency in transcription by using technology. What worked in improving efficiency was the transcriber's knowledge of medical terminology and medical field, matching the same transcriber with the same speaker as transcriber became acclimated with the nuances and tones of the speaker, and letting transcriber handle only certain type of transcription.
Transcription is a manual operation and Toyota Way is a good way to maximize efficiency.
Our goal with ASR is not replace the typists, but to assist them. On an average typing takes 4X the duration of actual time. An 80% accurate ASR may reduce it to 2X, which in turn should translate to faster turnaround and improved throughput.
See https://github.com/pannous/caffe-speech-recognition (still under development)
ProductHunt is going in the right direction - but unfortunately the products are already built with little or no market testing so I predict a high failure rate.
KickStarter is a nice way to see if there is some market demand, but also a high failure rate (55%) by people who lack the experience to execute.
Definitely you will need to have patience to find ones that catch your fancy. Also no one job going to give you idea, but when you read a lot of them, you start to see patterns and might develop linkage that will make interest project for you.
"He has an entire team of people who read reviews on Amazon, looking for moments when people say, "I wish this speaker were rechargeable." Pikarski then makes a rechargeable version. "
I did spend one Sunday afternoon reading up reviews on apple app store. while there were some pretty interesting ideas, I saw one huge issue - the search on app store is unreliable. One, there could be an app implementing the said "idea" or "feature" but somehow not readily visible. Second, no way to really dig and learn about your competitors. Third, the same search might hurt the app chances.
Here is a blogpost I wrote about them. http://www.davidmelamed.com/2013/11/22/what-customers-really...
Here is a Fast Company feature about them http://www.fastcompany.com/3021229/chaim-pikarski-the-amazon...
I know I would, thats why I asked the question that way :)