Randall Stross, the Digital Domain columnist for The New York Times, has been following Y Combinator, the Silicon Valley start-up incubator, for more than a year. His book, “The Launch Pad: Inside Y Combinator, Silicon Valley’s Most Exclusive School for Startups,” will be published next month by Portfolio/Penguin.

He got a rare insider look at all of the incubator’s fledgling start-ups in robotics, mobile apps, e-commerce services and software for companies. Here is his exhaustive, but not exhausting, special report for Bits on the current crop of companies Y Combinator has produced.

Peer through the window of a hospital nursery and you see rows and rows of cute babies. I think newborn start-ups possess the same kind of cheek-pinching appeal.

And nowhere can you find as large a collection of start-ups in a single place as those that come out of Y Combinator, the Mountain View, Calif., seed fund.

On Tuesday, YC, as insiders call it, staged its 15th semiannual Demo Day for investors. Sixty-three of the summer batch’s companies made a formal public presentation to investors and the press. That’s
a lot of newborns.

The Demo Day format hasn’t changed much as YC has grown over the years, attracting ever larger numbers of applications and funding larger batches of start-ups. The purpose of the exercise is to give investors
a tantalizing tidbit or two of information sufficient for determining whether to have the conversations that may eventually lead to an investment. Each presentation lasts just two or two and a half minutes, is filled
with astounding figures (“a $150 billion market!”), comes in thematically random order, and is stacked one on top of the other, in a blur. (Eleven more presentations came at the end that were designated
off-the-record.)

Sure, some of the babies appear to be a bit cuter than the others. But all deserve a chance to make their way. What can be done immediately is view the group portrait. YC’s summer 2012 batch captures prevailing
start-up ideas in one moment in time.

A little more than half of the start-ups that presented are attempting to sell to business customers. Let’s start, however, with the start-ups that are consumer oriented. Many are fledgling media businesses built
upon user-generated content. We users apparently have very short attention spans. The humor site 9gag , which serves up silly memes found by users — and draws 65 million unique users
a month — says “bite size is the future of entertainment.” Reelsurfer lets users reduce a video clip, which may already be short, to extremely short length.

For more conventional entertainment, there is Mars Frontier, a massively multiplayer, real-time strategy game, the first game produced by SpinPunch. What is not so conventional, at
least not yet, is the use of HTML5, which requires no download as Flash-based games do.

Tapin, which like many of its batchmates incorporates the twin motifs of mobile and social, is a mobile video app designed for a speedy start. Kamcord also is in
social and mobile entertainment: mobile game developers are invited to use its software so that gamers can record and share their gameplay. Another company makes the Coco Controller,
a smartphone case that has physical buttons for use with mobile games that are written to work with the case.

Circular applies the social theme to news, letting one’s friends act as a filter. Or, for those who would like to not have the social overlay on everything one does, there’s
Everyday.me, an iPhone app that “captures all your life moments” but whose default setting is private: not to be shared with others.

There is some playful experimentation with communication. Visitors to Voicegem’s Web site can record a voice message that is sent to an e-mail address rather than a phone number.
Hubchilla matches users up with strangers to have conversations via texting but keeps the participants’ phone numbers hidden. (One of its co-founders, who is a 20 Under 20 Thiel Fellow,
said he began coding at the age of 8.)

Tablets receive the focused attention of Grid, which makes a spreadsheet that does some tricks that no spreadsheet on a PC could ever match.

Amid the fascination with mobile, the Web has not disappeared completely, however. For sharing favorite images found on the Web, there’s Imgfave. For favorite links, with the pitch
addressed to office workers in particular, there’s Kippt. For finding stuff of our own that is stored in Google Docs, Collections provides
a “finder for the cloud.”

No fewer than four of the start-ups in the batch provide travel-related services. Flightfox runs contests in which travel agents and other humans compete to find the lowest fare for international
or multicity flights. Or travelers can use GetGoing, which describes itself as “the next-generation Priceline.” It supplies a deeply discounted fare for leisure travelers
who are willing to agree to go to either of two separate destinations — but cannot choose which one it is. Vayable offers unusual travel experiences, bringing travelers and
independent guides together. TomoGuides is preparing smartphone city guides that will be offered free.

Like Vayable, other start-ups are creating an online marketplace, bringing together suppliers and customers and offering prices lower than those available using existing middlemen in the physical world. Tastemaker does the same for interior designers and customers.

RegistryLove works with existing merchants but is attempting to create a consolidated wedding registry that spans all stores.

Healthy Labs, a start-up hosting what it calls “patient social networks” for chronic diseases, has begun with a site for those with Crohn’s disease or ulcerative
colitis: Crohnology.

In education, Knowmia has set up a site that invites teachers to contribute video lessons, what the company describes as “a crowdsourced Khan Academy.” StudyEdge offers college students study guides, video content and live review sessions that address challenging courses at particular universities. It’s now available at the University of Florida.

The personal finance category of software is an old one. SmartAsset, a Web-based service, applies its own analytical engine to the category, which supplies customized answers
to users’ questions like how much house can one afford.

Instacart is an intrepid start-up that has entered another old category, one known for the notorious failure of early pioneers: Web-based grocery ordering. Its founder stressed that
it carried no inventory (unlike the infamous failure Webvan): “We’re entirely software.”

A category of software with a much shorter history is that for 3-D printers. Dreamforge lets users create a customized design for a three-dimensional superhero that the company will print
— that is, manufacture — then send back to the customer.

The blend of the physical and the digital is also found in Vastrm, a Web site that offers customized shirts, including a “Home Try-On Program,” similar to what Warby Parker
offers for eyeglasses. Vastrm’s founder received one of the loudest appreciative laughs of the day when he pointed to Warby Parker’s success and remarked, “Not everyone wears glasses, but everyone
wears a shirt.”

Suppose you’re not home when U.P.S. tries to deliver your Dreamforge superhero or Vastrm polo shirt to your apartment? Someday you may live close to a BufferBox location, each
with a box containing multiple parcel lockers that are controlled through software. (BufferBox is rolling out its initial boxes in Toronto.)

The remaining start-ups in this summer batch are pitching to the enterprise or to fellow software developers.

Some services are focused on a particular industry and are highly specialized. Keychain Logistics brings together shippers and truckers, eliminating human brokers.
Viacycle runs bike-sharing programs for universities, cities and businesses. Rentio tells apartment owners what they should charge for rent. Eligible helps insurance companies and health care providers fully automate insurance eligibility queries, eliminating the expense of placing, and fielding, phone calls between humans. Hiptype teams up with e-book publishers, embedding its software in the books to gather data for what it calls “data-driven book publishing.”

If the business is one in which individuals outside the company send in documents that must then be reviewed — résumés, grant applications, college applications — there’s a YC company
that handles the process: Submittable.

Parallel Universe supplies a database service for businesses or organizations that manage vast amounts of spatial data in real time, like the Israeli air force, an early customer.

Amicus seeks to profit from helping nonprofits replace inefficient fund-raising using old means, like mail appeals and cold calls, with social media: friends referring friends.

Finance, with its large I.T. budgets, has attracted BigCalc, which does fast simulations for quants and algo traders, and DataNitro, which weds Python
financial and math libraries to Excel spreadsheets.

FundersClub wants to be “the future Nasdaq,” matching accredited investors with tech start-ups. Coinbase offers a way for nontechnical
users to get started with Bitcoin, the digital currency.

Many of the start-ups that are selling to the enterprise aim at customers that belong to no particular industry. MicroEval seeks to move performance reviews to the cloud. Statwing supplies easy “data visualization” to nontechnical people working with spreadsheets. QuicklyChat is a video chat service for co-workers that the company claims
is easy to use but will not interrupt the work of the person who is asked to respond because of the software’s “intelligent availability detection.”

Small and medium-size businesses that need a customized phone system can turn to Profig. For warehouse management, there’s HD Trade Services’
cloud-based inventory manager.

A large cluster of the start-ups are in marketing or advertising. Sponsorfied seeks to match corporate sponsors with targets. LeanMarket guides ad
buyers on banner-ad exchanges. Canopy Labs does customer modeling to help clients evaluate the value of leads. Referly expands affiliate marketing, working
with more than 30,000 retailers. Tracksby plans to arrange celebrity endorsements. Virool offers to help spread advertisers’ videos through social networks,
blogs and phones. On behalf of clients, Survata distributes market research surveys to Web publishers, who let visitors who complete the surveys read articles free instead of paying.

M^th Sense provides mobile advertisers with information about the gender and interests of users; the company deduces these by looking at what other apps are installed on a phone.
(This company holds what I believe must be the day’s record for largest figure mentioned — 25 trillion — a reference to the number of in-app ads that the company says will appear in 2016.)

We come to the last category of start-ups in the batch, the one that is, by far, the largest: tools and services for software developers. YC has always attracted strong technical teams and no one knows better than developers
and designers themselves what they need. This batch has Light Table, an interactive programming development that had raised more than $300,000 on Kickstarter last spring; Easel and Markupwand for Web site design; Scoutzie, a marketplace for mobile app designers; Airbrite for tightly
integrating e-commerce functionality to mobile apps; Clever, which has APIs for software developers who are working with schools and their student rosters; Plivo,
challenging the dominant incumbent, Twilio, which has APIs for voice and SMS applications; Filepicker.io, which eases file movement between
a local device and the cloud; and Zapier, which seeks to be the master translator of all APIs.

Representatives of one company in the batch stood out in particular; they were arguably the sensation of the day and had a thick knot of observers surrounding them at every break. These were the svelte robots of
Double Robotics, designed to provide telepresence inexpensively. One iPad is mounted on top, facing outward; another iPad is held by the human who controls the machine. In
developing it, the founders took on challenges in hardware as well as software, but they were able to come to market with a price of $2,000, a fraction of the price for telepresence robots from incumbents. The company
received $500,000 of advance orders in just one week.

Gratifying, yes, but being in the hardware business means dealing with real materials and finite supplies. The company sold out its first production run and says orders placed today won’t be filled until early
2013.