Data Science Consulting

We help companies build intelligent products.

We work with clients to build intelligent systems, using existing data. Our sprint-cycle process focuses on building custom testable technology for your data and goals. With each client, we work through the stages:
  • Data Deep Dive
  • Data Strategy and Product Goals
  • Prototyping and Solution Measurement
Some of our most successful clients are product teams who want to build, test, and realize the power of intelligent technology while aligning with their roadmap. We are not an R&D shop - we build testable technology utilizing already existing data and open source tools.

Data Strategy

From global corporate data strategy to startup metrics, we guide leadership on operation and strategy. We also advise companies on team structure, growth, and development processes.

Machine Learning

Whether it's modeling churn likelihood, which picture to show user segments, or projecting industrial trends, we model complex problems with machine learning.

Natural Language Processing

We've built scalable systems for information extraction, processing, and modeling natural language data. Our most recent custom scalable NLP system is now open source.

Featured Clients

What We've Achieved Together

"We hired Luminant to help build a quantitative report about the startup ecosystems in cities across the country. They were a pleasure to work with throughout the process: knowledgeable, thorough, engaged, and helpful. The combination of technical skills and general analytical ability made them more than a vendor building a specific product for us; it made them a true partner on the project, helping us to conceptualize new metrics to take our work to the next level and educate ourselves along the way. They are an ideal fit for anyone trying to tackle a complex data challenge."
Patrick McAnaney,

Innovation That Matters 2016

Wordnik is a dictionary that catalogs every word used in English on the internet. Our custom-built scalable Natural Language Processing and Machine Learning Model pipeline finds new definition of words. The final product, Serapis, is an open-source end-to-end solution using Amazon Lambda as a task pipeline.

Scouring the Web to Make New Words ‘Lookupable’ The New York Times

Serapis on Github

Development of Python tutorial series for business analysis using real-world data.

Python Tutorials