Rohit Kelkar

Rohit Kelkar is a data scientist with 18 years of experience researching, developing and launching artificial intelligence and machine learning solutions for legal and compliance. As a senior member of FTI Technology’s E-Discovery Innovation team, Mr. Kelkar leads ongoing testing and proprietary workflow development for IQ.AI by FTI Technology, AI offerings for disputes and investigations, antitrust and data breach response.

Mr. Kelkar is an expert at defining technology roadmaps in alignment with client goals and managing cross-functional teams to execute against technology plans. He has managed large teams of data scientists, software engineers and machine learning operations to deliver impactful products and innovations. His work has also included evaluating, testing and customizing a wide range of large language models for legal tasks.

Mr. Kelkar helps clients navigate complex challenges across the legal domain, including in e-discovery and complex investigations and regulatory matters. He is highly focused on upholding defensibility and tailoring solutions to meet the specific needs of each case.

Prior to joining FTI Technology, Mr. Kelkar served as Head of AI/ML Products at a software-as-a-service company, providing AI-driven technology solutions to legal and compliance departments and large law firms. In that role, he collaborated across engineering and business teams to identify, scope and execute AI projects, build proofs of concept for clients and explore new research and development opportunities.

Relevant Experience:

• Led development of a time series machine learning algorithm to predict drill failures, resulting in reduction of drill failures; also architected a bid processing platform using cloud architecture to reduce operational costs as a consultant for the data science and analytics group at a large energy company
• Mentored a team of data scientists to devise scalable large language model algorithms for analyzing large contracts, cutting down human review and saving costs; also established a large language model test harness to reduce model experimentation and development time and increase productivity