AI Solution Development

Artificial Intelligence (AI) is leading the charge in the new cycle of economic change “Fourth Industrial Revolution”. While the impending changes hold great promise, the recent breakthroughs in this technology also pose major challenges requiring proactive adaptation by enterprises around the world.

Companies aspiring to be successful in this new economic landscape must harness the competitive edge of technological change and hire the best practitioners, avoiding being left behind by perilous talent shortage. Investments in reskilling and upskilling today’s workers will be critical.

The net effect of AI in the supply of skilled labor is most businesses currently face major recruitment challenges and talent shortages.

Centers of Excellence provide a comprehensive solution for companies to successfully transition into being AI-centric by offering enterprises a program to cultivate world-class AI talent.

A great data science team is one able to solve data-driven problems/challenges regardless of prior domain expertise in the field in which the problem arises
— Jeremy Howard, Founder Fast.AI

Centers of Excellence
de-risk AI

There are several reasons why organizations implement centers of excellence to remain successful in competitive markets. The definition of a CoE states that they are a group of people leading the organization and its different structures in a specific focus area towards pre-determined goals. Hence, the aim of a CoE is to improve expertise in a certain area and make the most of its resources to help the business to improve.

Unique Methodology

We  collaborate with industry partners to explore and/or broaden the use of ML in their organization. Launchpad.AI delivers high-value machine learning solutions for enterprises by using the agile software-development principles and applying cutting edge deep learning and large-scale optimization techniques to computer vision, natural language processing and time-series problems. This allows us to run an efficient process for building advanced machine learning models that are ready to be used in production by our partners.

  • Projects are scoped to deliverables that a dedicated team of partners  can tackle in 3-6 months

  • We use the latest machine learning tools and techniques, like for example: self-supervised and multi-task deep learning methods, online semi-supervised labeling methodologies, transfer learning from benchmark pre-trained models, wide and deep neural networks that support sparse input features, and large-scale optimization of models and processes

  • We establish weekly checkpoints with a status update presentation and discussion about next steps

  • With agile software development methodology we include weekly iteration planning, daily scrum and story driven development

  • We guarantee complete visibility and transparency (shared Github repos, Trello boards & Slack groups)

  • Building the right environment for managing machine learning experiments (

Team Composition

  • Channeling the enthusiasm to learn into real project deliverables (fellows)

  • AI practitioner with a proven track record for delivering ML projects in industry (Launchpad principal / tech lead)

  • Leveraging the experience and knowledge of our team of AI experts who stay current on the most advanced AI techniques (research advisor)


Operationally, the technological direction of the CoE is very much in line with Francois Chollet’s view on the areas of most promise in the most advanced fields in AI:

  • Models closer to general-purpose computer programs, built on top of far richer primitives than our current differentiable layers—this is how we will get to reasoning and abstraction, the fundamental weakness of current models.


  • New forms of learning that make the above possible—allowing models to move away from just differentiable transforms.

CoE Roadmap

We work with industry partners to apply machine learning to high-value problems. We ensure that partners derive tangible value from applied AI through our unique methodology, suite of tools, and access to prominent researchers and open source contributors. We have successfully delivered AI solutions for leading hedge funds, some of the largest telecom operators in the world and emerging fintech startups. Multibillion dollar hedge funds currently trade with systematic strategies developed by our team.


Successful CoEs are built around a specific focus area

This should be an area relevant for the company and of particular importance for the business.



  • 4 full-time participants nominated by company (vetted by Launchpad.AI)

  • Experienced machine learning practitioner from Launchpad to lead the project

  • Define Proof-of-Concept (PoC) project and deliverables

  • Team augmented with fellows (if necessary)


  • Push PoC project(s) into production

  • Quantify and advocate results from PoC

  • Establish machine learning experiment management toolset

  • Plan second cohort (10 participants)

  • Appointing research advisors based on appropriate ML sub-field


  • Transition fellows to internal team

  • Support ML tools

  • Advise on machine learning projects

  • Assist in identifying and assessing AI opportunities

Fellowship Foundation

Based on Fellowship.AI’s proven framework, Launchpad.AI establishes and operates Centers of Excellence in Artificial Intelligence inside enterprises. Instead of competing for scarce and expensive resources externally, the CoE program enables enterprises to (1) develop highly proficient machine learning engineers from their existing talent pool; (2) establish the required environment, tools and best practices for consistently developing AI solutions; and (3) establish a recruitment engine inside their organization to continue growing their teams by attracting, retaining  and developing world-class talent.

  • 60+ fellows have graduated since 2015
  • Approximately 50% of applicants have a PhD
  • Less than 6% of all applicants are accepted

From Fellowship to Launchpad

A select group of fellows that demonstrate extraordinary ability or achievement are recruited to work on commercial Launchpad engagements after completing the fellowship.  

This allows us to industrialize exploratory projects started in the fellowship with team continuity.   


70% of fellows continue to collaborate with the program in some capacity. The majority of those who continue engaged with the program are now mentors to the program’s most recent cohorts.

A select group of fellows that demonstrate extraordinary abilities are recruited 


Past Fellows

Successful outcomes and high satisfaction hires by: Uber, Facebook, Enlitic, Sentient Technologies, Ernst & Young, Engineers Gate, Yelp, Orange, Pivotal, Lumiata, 6sense, SF Motors and Allstate