Contrary to popular belief, we think that the rise of Artificial Intelligence will be a boon for small-and-medium enterprises in the B2B space. For the foreseeable future, business applications of AI require significant human intelligence in the form of design, programming and supervision. AI pioneers like Google and IBM are opening up access to their technologies as a way to get businesses to leverage in their platforms and create new services.

The movie “Hidden Figures” has an interesting story arc about Dorothy Vaughn re-training her team of human-computers to become FORTRAN programmers. In the movie, she tells a skeptical team-mate: “[The IBM mainframe] will run eventually. We have to know how to program it once it does... Unless you’d rather be out of a job?”

Artificial Intelligence is in a similar stage of development as mainframes were in the 1960s. We need to start recognizing the power and limitations of Artificial Intelligence to incorporate it into business processes.

LambdaZen has worked on several technologies related to what is broadly referred to as “Artificial Intelligence” including:

1. Speech Recognition: Google, Amazon, Microsoft and IBM have all released very powerful speech recognition APIs that transcribe audio to text in online and batch modes. We have hands-on experience with these APIs.
2. Computer Vision: Again, the big players have publicly accessible APIs for computer vision. These APIs extract text, detect faces and classify images. The face detection and classification algorithms can also be trained with pre-existing data sets. We have integrated some of these facial detection and text extraction APIs in our products.
3. Natural Language Processing: NLP focuses on extracting syntax, semantics and sentiment from plain text sentences. We have deep expertise in sentiment analysis and lead the development of an industry-leading solution in this space.
4. Rule-based expert systems: PROLOG was one of the first technologies to be labeled as Artificial Intelligence. It has been used to build rule-based expert systems for various use-cases such as threat detection and medical diagnosis. We have developed an open-source PROLOG engine for graph databases that can efficiently perform sophisticated traversals on Connected Data based on a rule-based domain model.
5. Machine Learning: These are powerful pattern recognition algorithms that can be trained for specific applications. We have worked with clustering algorithms, Bayesian networks and Artificial Neural Networks, which are popular approaches to machine learning. One such clustering algorithm developed by our team has been deployed on IoT devices at thousands of retail locations.

Can AI supercharge your business? We’d love to help!