The Series A round, led by Physic Ventures, is its first since Intel Capital, Harrison Metal Capital, First Round Capital and angel investors forked over $1.85 million in seed funding in 2010. All of Yummly’s original investors stuck around for another injection and were joined by Unilever Corporate Ventures and Harvard Common Press as well as Physic.

Feller and CTO (you guessed it, Chief Tasting Officer) Vadim Geshel used that seed funding to build Yummly’s core technology, which breaks down recipes and ingredients into raw culinary data, from which it extracts everything from calorie and saturated fat content to more conceptual notions of saltiness, sourness or bitterness. If it knows you’re a vegetarian, the Yummly search engine will weed out recipes that contain an ingredient that once walked, flew or swam. If it knows you like spicy foods, it will steer you to the recipes of pepper-loving regions of the world, rather than those of Northern Europe. Or if you’re on a diet, it can help you plan a meal down to the nearest carb or calorie.

But Feller said that the semantic technology Yummly has built has a lot more potential than the search engine lets on. Its current staff of 10 just doesn’t have the resources to build new products to tap into that potential. So Yummly plans to use its new $6 million to go on hiring spree and roll out new cooking applications and features. Some of those products will be the typical tablet apps and shopping list services we see proliferating on cooking sites and services today, but others will make more specific use of the vast food vocabulary Yummly has spent two years learning, Feller said.

“For instance, we’re able to understand when a recipe requires you to poach an egg,”  Feller said. “We can then push a video to you that shows just how to poach an egg.”

That may not sound like a big deal, but in the world of today’s Web, “poaching” is just a searchable term, not a series of cooking techniques that vary depending on the ingredients used or the style of cuisine employed. Poaching an egg in acidulated water is vastly different than poaching salmon in butter, just as caramelizing an onion uses a technique altogether different from caramelizing sugar. If Yummly can ferret out that context from raw recipe data then it will have accomplished something big.

First things first

Feller said Yummly first needs to build upon its core search engine, which he characterized as being more a proof-of-concept –- though its a proof-of-concept that gets 4 million unique visitors a month. It currently aggregates more than 600,000 recipes from 15 sites, ranging form the Food Network to AllRecipes, but it will expand its scope into further culinary corners of Web. It plans to add personalization features that will make Yummly more of a cooking portal than a search engine. And it has to find a revenue stream.

Ads have started showing up in Yummly searches, but Feller said the company has only been experimenting with them so far. Feller, however, has little doubt that Yummly can build a solid advertising business once it gets serious. The data it collects will make ad targeting far for refined than anything you see on AdWords.

“Say I’m an advertiser that wants to target vegetarians that like quinoa,” Feller said. “That’s an even more powerful concept when you consider that 10 percent of the GDP is food.”

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Social media in Q1: commerce and discovery dominated
• A near-term outlook for big data
• The capex connection: Why we pay for privacy on the Web

Users can sign in to Beepl with their Facebook, Twitter and LinkedIn accounts, linking the new service to their social graphs. Once a user has logged in, Beepl scans their social media profiles to create an interest graph. The company uses semantic technology, natural language processing and machine learning to build its network of subject matter experts and to deliver relevant questions to them.

The idea is that “users don’t follow topics, but topics follow them.” Beepl also ranks users, creating an algorithmically determined “Beepl score” based on their contributions to the site. That helps it further identify users who provide valuable input on the site.

While using people’s existing social graph is a good, fast way to determine what they are knowledgeable in, linking to their Twitter accounts also enables Beepl users to send questions to the system using the #ask hashtag. In addition to responses that happen on the site, Beepl will also pull in any responses that emerge from their Twitter followers.

Beepl, which is headquartered in London but has an office in Prague, was founded by CEO (and former TechCrunch journalist) Steve O’Hear. The startup has spent the past three months in a private alpha period before preparing for launch. In July, Beepl announced a seed investment from Credo Ventures, and Credo partner Ondrej Bartos sits on Beepl’s board.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Social media in Q1: commerce and discovery dominated
• Startup growth and the new recruiting ecosystem
• Pinterest reawakens Napster-style debate over copyright