Senior manager - data scientist

IBM Research, Yorktown, NY                                                                    2008 – July 2013
Manager, Intelligent Information Interaction

Manage teams focused on different aspects of intelligent information interaction, including:

  • Smarter Decisions (2008 – 2009) Creating visualizations and a framework for decision trails, which helps organizations make smarter decisions by automatically capturing trails of the analytic process. Trails can be bookmarked, shared, and restored to understand the provenance of such decisions, as well as to facilitate collaborative decision making and expertise transfer.
  • Next-gen Meetings (2009 – 2011) Developed a web-based solution that enables users to upload recorded meetings, automatically create transcriptions and attach tags and comments. Supports surfacing information buried in monolith recordings by making it accessible from a collaborative point of view, as well as from an information mining aspect. Metadata is used to facilitate searching for segments of interest, as well as collaboration and discussion.
  • Lead for the Consumability of Analytics Strategic Initiative, and the IBM Center for Advanced Visualization

Sample Projects:


Jennifer Lai is returning to NY after a two year assignment at theIBM Research lab in Australiawhere as a Senior Manager she is responsible for the Cognitive Computing mission. She leads more than 20 researchers and software developers who are driving innovation into the next generation of solutions in real-time social analytics, as well as image and video analytics. An accomplished data scientist and leader with over 25 years of experience working in R&D, she has a proven track record of leading projects of an innovative nature with high impact. While on assignment in Australia, Jennifer defined a new area of research for the lab creating whitespace for IBM: real-time, multi-channel social analytics for cognitive computing and is often invited to speak on this topic, with 6 keynote invitations since arriving in Melbourne.
Jennifer’s expertise is in the interdisciplinary area of deriving insights from data (unstructured, structured, or multimedia) with intelligent user systems, including interactive visual analytics and people-centric designs. She manages multiple teams of researchers and engineers whose skills range from machine learning and NLP, to visualizing data, to analyzing video and images and as well developing cloud-based services for clients. As the creator and leader of the IBM THINKLab in Melbourne, Jennifer got the space transformed from cubicles to an open client space for innovation and discovery. Touted as the model for THINKLabs in all global labs, in the first year the lab hosted 27 account visits, which exposed more than 103 key clients to the leadership of IBM Research.
With 23 patents issued, several patents pending and over 56 publications Jennifer is an internationally known scientist whose work has been recognized with multiple Outstanding Technical Achievement Awards and three IBM Research Accomplishments.  Her publications have been cited 3,203 times with 1,364 citations since 2010. In addition to management, Jennifer has deep technical expertise in social media, experience design, cognitive computing and multimodal technologies, including speech and image recognition. Media trained, she has been interviewed by all the major media channels (e.g. NPR, NY Times, WSJ) on multiple occasions for her work.

IBM Research, Hawthorne, NY                                                                        1987  – September 2004
Research Staff Member - Multimodal Technologies


  • 1987 –  1996  Helped define the field of Statistical Machine Translation (SMT) with Bob Mercer and Peter Brown (now CEO of Renaissance Technologies) by working on aligning the Hansard Corpora. As such I was part of the team that wrote the core initial patents on SMT. Later, I focused more on Language Model development, as well as tools to automate the process.
  • 1997 – 2000 Defined and established most of the guidelines for designing effective conversational speech systems. Given the highly variable nature of speech (accents, microphone issues, background noise), we do not expect consistently perfect accuracy from automatic speech recognition in the foreseeable future. Effective design for spoken interfaces requires concentrating on error prevention by creating prompts that cue the user what to say, providing seamless error recovery, and using techniques to speed the interaction. 
  • 2000 –  2002 Measured and published human comprehension of Synthetic Speech (also known as text-to-speech TTS), in various interfaces including automotive. When I started using TTS extensively in applications, there were virtually no studies that looked at general comprehension. Two key studies I published indicated that listening to low quality synthetic speech appears to interfere with short-term memory processing, and that using TTS exclusively for system prompts in an interface is preferable to mixing synthetic speech with recorded human speech.
  • 2002 –  2004 Created a next-gen workspace, the IBM BlueSpace a prototype ‘office of the future’ installed at the IBM Industry Solution Lab in Zurich and New York. With a focus on increasing productivity by deterring unwanted interruptions, improving awareness and fluid communication among team members, and providing greater individual comfort through personalized environmental settings. BlueSpace was featured in all major media avenues (e.g. New York Times, Bloomberg, Associated Press, International Herald Tribune and USA Today) television  and more than 12 different US regional newspapers, 10 magazines and numerous on-line news sites.


IBM Research, Melbourne, Australia                                                             July 2013 – Present
Senior Manager, Cognitive Analytics                                                                                                                                                                                                                             
Key Responsibilities:

  • Driving innovation into the next generation of solutions in real-time social analytics, as well as text, image and video analytics;
  • Defining and leading the Cognitive Analytics research mission in Australia including helping Watson “understand” breaking news from social media, and teaching Watson to “see”. 
  • Managing up to 10 different projects and multiple teams with a diverse set of skills including machine learning, NLP, data visualization, multimedia analytics as well developing cloud-based services for the market;
  • Client interactions including Proof of Concept (PoC) definition and delivery, as well as creating a vision for innovation based on an understanding of core business needs;
  • Building bridges between data scientists and client teams by establishing common goals fostered by mutual understanding;
  • Leading cutting-edge prototype solutions to delivery using an iterative process and transferring the learnings to products or clients, while recovering costs for the research team;
  • Achieving balance between the need for real-world business impact and the need for deep and meaningful science that results in patents and publications;
  • Successful hiring and growth of teams;

 Sample Projects:

IBM Research, Hawthorne, NY                                                                   September 2004 –  2008
Manager, Adaptive Learning  

Manage team focused on different aspects of adaptive learning, including:

  • Web-based Literacy (2004 – 2006):  Developing and delivering a web based  interactive learning system for young readers (grades K – 2) and adults (primarily an immigrant population) using engaging UIs with an animated tutor and multimodal technologies including speech recognition and text-to-speech;
  • Synchronous expertise sharing  (2006 – 2008) Expanding expertise within the organization with a real-time expertise sharing and capture application to connect information-seekers to people with the expertise to answer them, while placing safeguards around the experts’ time.  Tool was widely deployed throughout IBM to help with the need for seamless expertise transfer between different global service delivery centers.                                                                                                                                      

Sample Projects:

Jennifer C Lai

jlai324(removethis) or jelai[delete] at


  • Selected as one of the “top 10 innovators with Twitter at IBM”; press release August 2015;
  •  27 patents filed, including 23 patents issued;
  •  Over 56 papers published in top conferences, journals, book chapters and encyclopedias;
  •  Media trained and interviewed by many of the major media channels including NPR, NY Times, WSJ, Business Week, CBS TV, CNN, Bloomberg (e.g. Harvard Business Review ( ), Discovery Channel  (“The most extreme – Workaholics”)
  • Invited lectures in academia include: MIT University, New York University, University of Maryland (UMBC), Michigan State University, Pace University, University of Tampere, Finland, Center of Advanced Judicial Studies, Royal Melbourne IT University, Melbourne
  • Keynote speaker at conferences include: User Interface Software Technology (UIST) Conference;  Visualized; Australasian Language Technology Workshop;  Efficient Consumer Response Australasia; IBM Connect; IBM Impact;
  • Work featured in ACM Digital Library: “Superhuman speech by 2010: an interview with Jennifer Lai” and featured scientist in ACM Crossroads: “A day in the life of...Jennifer Lai”;
  • Guest editor for Communications of the ACM, and International Journal of Speech Technology;
  • National Science Foundation Panelist;
  • Member Editorial Board - International Journal of Speech Technology