TRACK BILLIONS OF CONSUMER DECISIONS Real-Time Insights from Millions of Current and Potential Users

Contact if you want to know about DRUG_NAME.

Contact to upgrade your subscription.

Please contact to gain access to DRUG_NAME.


DecisionEngine is...


Connecting you to millions of customers. Anywhere. Anytime. Any device.


2.5B unbiased reports. 50M healthcare customers. Whoa. Find the right ones.

Ready to Engage

Online conversations drive consumer decisions. Capture consumers in the moment of decision.

Optimize Marketing & Substantiate Investments

DecisionEngine tracks and predicts purchase intent of every healthcare brand. To do that, we analyze the transient needs and decisions of millions of people talking about their health online. Our repository of patient insights is the world's largest.

We are more than the sum of things we click. We recommend. We complain. We hesitate. We decide and take action. With over 80% of consumer decisions starting online (McKinsey, 2012), these decisions can make or break your brand.

Demystify Your Consumers' Rationale

Over 95% of the world's public data is unstructured - not analyzed by anyone, in any depth. Brand strategists can't access this data. Neither can media planners. Your current analytics solution tracks words, sentiment, and themes, not decisions, motivations, and intent. Your ad platform tracks visits, not meaning. Traditional advertising metrics - views, clicks, signups, sentiment - can't answer even your basic questions:

- Did my campaign impact consideration of my product?

- Why do people choose my competitors?

Next time you are told "sentiment about your brand is up", ask "what does that mean?". Avoid costly mistakes resulting from analyzing topics and sentiment. Base your decisions on accurate, representative, and trackable consumer decisions.

Find Critical Brand Insights

Online conversations are filled with incredible promise - if you have the right map to its treasures. Finding answers requires a good net - wide enough to capture all the relevant data, and selective enough to filter out all the trash.

Wielding the top scientists in text analytics (MIT PhDs), we have been building such a net - the DecisionEngine. We are teaching computers how language works, how to think like we think, how to take apart meaning of conversations, how to eliminate the Internet noise: the retweets, the fake sites, the aggregators, the paid bloggers.

DecisionEngine automatically zeros in on the right data, with the critical precision you need to understand your customers and fend off the competition.

Goodbye Miss Sentiment. Hello Lady Action.

Getting insights out of social media is incredibly hard. Your analysts are not at fault - the tools are. Counting words, analyzing sentiment, and clustering topics are exceptionally poor means to structure data.

Analysts are forced to dig through piles of irrelevant data and create insight from a tiny, unrepresentative fraction of public chatter they had time to find, read, and structure.

DecisionEngine eliminates the need to sift through public data. It answers your most pressing questions: why consumers act, what they plan to do, and how to reach them in their moment of decision.

Intervene Preemptively

Accurately predicting consumer decisions has a direct impact on your media spend. Why pay for impressions and clicks when you can connect with most likely consumers, at their moment of decision? Automatically support consumers that endorse your product; offer an alternative to consumers dissatisfied with your competition. DecisionEngine is tailored to position your brand in front of most likely converts, in a manner that is both regulation-safe and easily measurable.

Social opportunities (and risks) emerge in the space of days, sometimes hours or minutes. Can your company wait?

  • DecisionEngine gets to the key social insight - why consumers act - and helps you take action.


Choose what fits your needs...


Access the best-looking dashboard on the market (our clients' mothers say so). Pay per:

  • Number of brands you track
  • Features you need (details here)
  • Number of seats


Ask any question. About any brand.
Power users' nirvana.

  • DPC-based pricing (Data Processing Credit). More use = lower DPC cost (details).
  • $250 startup fee (waived once you reach 100,000 DPCs).
  • See the exact cost of each query before running it (details). No surprises.

Custom Work

Ask any question about consumer conversations, in any industry. Examples:

  • 1. Justify regulatory approval. Public poster.
  • 2. Compute ROI on a TV+Social campaign. Combine social and proprietary data (report excerpt).
  • 3. Position product for entry to the US market (competitive weaknesses analysis excerpt).
  • 4. Create a litigation map of the USA. Web service analyzing unstructured portions of a live stream of SEC filings.
  • 5. Analyze 50,000 open-ended responses to a focus group questionnaire. Proprietary data analysis.Public poster.
  • 6. Correlate social insights with 3rd party data. Integrations include Earnings calls, news feeds, BLS, FDA, USPTO, etc.
  • 7. Monitor user submissions to public web properties owned by pharma companies. Ensure compliance with regulatory reporting while eliminating human review. Deployments demonstrate 97% correlation with human review.

  • Understand, Intervene, Shape.

    Real world. Real time. Real accurate.


Relationship Graph for Unstructured Public Data (98% of it all)

Founded by two MIT PhDs, our team includes world authorities in natural language processing. Our work has been cited in over 4,500 academic publications. Our innovations have received five consecutive National Science Foundation awards.

Our clients include top pharmaceutical and healthcare companies, as well as hedge funds, ad agencies, and publishers. Luminaries including Nicholas Negroponte (founder of the MIT Media Lab), David Lee, MD PhD (Founding Executive of Endo Pharmaceuticals), and Declan Doogan MD (Former Head of Worldwide Development at Pfizer) serve as company advisors.

Our goals:
Identify every consumer and every potential consumer for every brand.
Connect every Internet user to other users that have similar problems and useful solutions.
Create the first conversation-driven, unbiased scorecard for every brand and company.


Ritwik Banerjee


Ofer Bronstein


Xavier Carreras


Zurab Khukhashvili


Gabriele Musillo


Edward Nemirovsky


Paul Nemirovsky


Joshua Pevner


Audi Primadhanty


Ariadna Quattoni


Sophia Van Valkenburg


Deanna Wallach


Come work with us!


Be part of a team revolutionizing analysis of unstructured text and consumer knowledge.

Work closely with others to build a solid product stack. We believe in giving each employee ownership over an important part of our technology, product, and service.

Explore large-scale content aggregation, data processing, and storage. Provide leadership, code discipline, and project design by example.


Front End

Engineer sleek solutions for visualization, presentation, and interaction with high volumes of semantically-rich longitudinal business intelligence data.

Develop beautiful and easy to use web applications that render a consistent output.

Your users will range from Fortune 100 to individuals using their iPhones. Educate us on the possibilities. We do not need to support older browsers, so you are free to unleash your creative knowledge.

Passionate about UX. Loves the challenge of designing elegant UIs for non-tech savvy users. Creative with both code & markup.

Experience with: Microframeworks such as flask or; or Django (X)HTML, HTML5, CSS, JavaScript, cross-browser techniques iOS, GWT (bonus).

Back End

Use your expert Java skills to solve a wide variety of engineering challenges, ranging from data flow, to storage, to aggregation, to supporting APIs and the presentation layer.

Design and build high-load web applications and service-oriented systems for storing, processing, and searching a very large volume of unstructured text.

Implement a SaaS information delivery architecture. Deploy to and manage apps in the cloud.

BONUS: Knowledge of machine learning, statistics, natural language processing. Designing and consuming RESTful web service APIs. Hands-on experience with Hadoop / MPI or equivalent. Experience with cloud deployment. Experience with Python, Ruby, bash/unix tools.

CS Interns

Want to get experience while avoiding the corporate monkey position? We have a few projects that would love your attention, in data exploration, analytics, annotation, and more



We take great pride in running our machine learning and NLP algorithms on real world datasets at real world speeds. A successful candidate will: Help researchers implement and optimize their algorithms for near real-time performance over terabytes of data, and integrate these algorithms into our overall NLP pipeline. Work heavily with our application development team and coordinate research work with product development. Have a Masters in CS (or equivalent), with experience in distributed systems and/or high-performance algorithms.

Data Mining

Validating pieces of information extracted from our data is business-critical. Research work in data mining will focus on assessing both the predictive power and the novelty of extracted information. In addition to validating extracted information, the ideal candidate will provide feedback to the natural language processing research team as to which pieces of information to extract in order to gain further statistical insights.

Semantic Parsing

Mapping textual data to representations of meaning suitable for data aggregation is one of our core objectives. The ideal candidate will design broad-coverage semantic representations that are deep enough to capturing linguistic phenomena relevant to data aggregation and build statistical models to learn these representations given limited supervision.

Data Annotation

Achieving consistent annotation needed for training target statistical models is one of our core research tasks. The ideal candidate will implement annotation procedures, possibly relying on crowd or community sourcing. Additional tasks include curating linguistic resources relevant to information extraction tasks. A background in linguistic analysis of real-world text data is required. Working knowledge of NLP tools is a plus.


To find us...


By bicycle

  • 181 North 11th Street
  • Brooklyn, NY 11211