Machine Learning & AI
The benefits of machine learning are material, but there are hazards to avoid.
While many machine learning projects produce orders-of-magnitude improvements, others stumble. Often, algorithms are badly built, not fit for purpose, or their data needs are unsustainable. Even when models are accurate, enterprises don’t always successfully integrate them into their production workflows. For executives, the key question is how to get started safely and cost-effectively.
Machine learning use case frequency in B2C environments*
*Algorithmia, 2020 State of Enterprise Machine Learning, Nov. 2019
Our approach de-risks the machine learning journey
Our process focuses on the viability of various machine learning (ML) solutions and the economic impact to the business. Two checkpoints along the journey ensure there is a return on spending before committing to production development and deployment.
1. Opportunity Assessment
- Where are the opportunities?
- What are the financial impacts?
- What processes would change?
- What are the systems impacts?
- What are the potential ROIs?
Checkpoint: Are ML opportunities feasible and with sufficient ROI?
2. Modeling Proofs of Concept
- Where are the opportunities?
- What are the financial impacts?
- What processes would change?
3. Production Architecture
- What are the systems impacts?
- What are the potential ROIs?
Checkpoint: Are models robust at scale, and is there still return on investment?
4. Production Implementation
- Development of scalable production ML models
- Integration and testing
- Data pipeline development
- Incubation of production workflows & training
- Iterative improvements during incubation
- Full rollout
5. Ongoing Production Refinement
- Monitor models for ongoing accuracy
- Refine models and processes
- Expand first uses to other uses
Meet a few of our experts
CLAUDIO CALVINO, Ph.D.
Managing Director
Claudio Calvino specializes in the development of advanced statistical methodologies and machine learning algorithms. He has designed and developed supervised and unsupervised fraud detection solutions and led development of advanced natural language and geo-spatial algorithms for clients.
RICHARD PALMER, Ph.D.
Senior Managing Director
Richard Palmer works with clients to use sophisticated technology to address complex legal, regulatory and business matters. He brings deep domain expertise in developing bespoke analytics solutions through the entire lifecycle.
YUVAL SABAN
Senior Managing Director
Yuval Saban has advised clients on the applicable and beneficial use of machine learning and other AI-related methods. At IBM he led large-scale cognitive engagement in U.S. banking, and previously was the co-founder of Deloitte’s Cognitive & Analytics practice.
ANNA ARLORIO
Senior Managing Director
Anna Arlorio advises clients on driving business value using machine learning and AI techniques. She has developed proprietary methodologies applied to the full spectrum of C-level business problems, from marketing to customer experience, sales and pricing, as well as the technical deployment of these solutions.
ETHAN WILANSKY
Managing Director
Ethan Wilansky is a full-stack cloud architect who has developed commercial solutions in both Azure and AWS. He is the author of 14 published computer books for Microsoft Corporation, and has over 100 published technical articles and holds various U.S. patents.
MELORIA MESCHI, Ph.D.
Senior Managing Director
Meloria Meschi has double doctorates in economic policy and applied econometrics. She specializes in applying economic and statistical analysis to complex regulatory, policy and strategy issues, including the measurement of causal impacts and the outcomes of machine learning algorithms in terms of fairness, discrimination and bias.
Other advanced analytics services
FTI Consulting provides a broad array of other advanced analytic services across 16 industries, including healthcare, telecom, media, transportation, industrial manufacturing and banking.
Retained data scientist services
ML/AI as an FTI-managed service
Design-build-operate-transfer (DBOT)
Business analytics, dashboards & visualization
Automated fraud/compliance detection
Unstructured data analytics at scale
RPA prototyping & development
Geo-spatial analytics
Predictive analytics
e-discovery
Forensic data analysis
Enterprise data management
