Neha Sengupta

AI Researcher

United Arab Emirates14 yrs 9 mos experience
Most Likely To SwitchHighly Stable

Key Highlights

  • Expert in Arabic NLP and bilingual LLM development.
  • Strong algorithmic background from PhD thesis.
  • Proven track record in experimental research and algorithm development.
Stackforce AI infers this person is a specialist in NLP and algorithm development within the AI and data science sectors.

Contact

Skills

Core Skills

Machine LearningAlgorithms

Other Skills

Applied SciencesApplied TechnologyCC++Core JavaData StructuresDatabasesEclipseGraph TheoryJavaJava Enterprise EditionJavaScriptLarge Language Models (LLM)LinuxMatlab

About

My professional experiences range from research to development, with recent focus on Arabic NLP and LLMs. I have been central to the development of Jais, a series of Arabic centric bilingual LLMs. My PhD thesis focused on algorithms in evolving graph structures, giving me a strong algorithmic background. I am passionate about experimental research, and developing new algorithms and systems.

Experience

G42

4 roles

Principal Applied Scientist

Promoted

Jan 2022Present · 4 yrs 2 mos

Lead Applied Scientist

Promoted

Mar 2021Jan 2022 · 10 mos

Applied Scientist

Promoted

Jun 2020Mar 2021 · 9 mos

Applied TechnologyMachine Learning

Research Associate

Sep 2018Jun 2020 · 1 yr 9 mos

Indian institute of technology, delhi

PHD

Jul 2014Sep 2018 · 4 yrs 2 mos · Greater Delhi Area

  • Publications
  • 1. [Full Paper] N Sengupta, A Bagchi, S Bedathur, M Ramanath, “Sampling and Reconstruction Using Bloom Filters” IEEE Transactions on Knowledge and Data Engineering (TKDE) 2017
  • 2. [Full Paper] N Sengupta, M Hamann, D Wagner, “Benchmark Generator for Dynamic Overlapping Communities in Networks”. IEEE International Conference on Data Mining (ICDM) 2017 New Orleans, Louisiana, USA
  • 3. [Short Paper] N Sengupta, A Bagchi, S Bedathur, M Ramanath, “Sampling and Reconstruction Using Bloom Filters”. IEEE International Conference on Data Engineering (ICDE) 2017. San Diego, California, USA
  • Projects
  • Sampling and Reconstruction Using Bloom Filters : Developed a novel data structure and algorithm to sample an element from a set stored in a compacted data structure.
  • Approximating Reachability in Dynamic and Versioned Graph Data : Designed algorithms based on random walks on a graph to answer the reachability query.
  • Internships
  • IBM Research, India(May 2017 - Jul 2017)
  • Extending the Traversal API of the TInkerPop3 framework to support time-based queries on a versioned graph
  • Karlsruhe Institute of Technology, Germany (Jul 2016 - Dec 2016)
  • Developed a dynamic graph generator that maintains statistical properties like those of real world graphs while simulating events such as emergence and dissolution of communities
  • IBM Research, India (May 2016 - Jul 2016)
  • Used IBM Genesis, that leverages the benefits of Apache Ignite for efficient processing in massive graphs on a distributed system, to implement a subgraph stream ingestion system for a versioned graph
  • IBM Research, India (May 2015 - Jul 2015)
  • Chronological Graphs on Spark: Maintenance of massive dynamic graphs in memory using Apache Spark’s GraphX platform. Worked on efficient ingestion of rapidly incoming data and fast querying of Time Varying Graph

Ibm research india

Software Engineer

Jul 2011Jul 2014 · 3 yrs · Greater Hyderabad Area

  • Received IBM Certificate of Appreciation for contribution to the Microgrid Management System
  • Research Areas
  • Data Mining for Smart Grid Applications
  • Renewable resources, storage – integration and optimization
  • Publications
  • 1. [Full Paper] M Jain, H Khadilkar, N Sengupta, Z Charbiwala, K U Tennakoon, R Wahab, L De Silva, DP Seetharam, “Collaborative energy conservation in a microgrid” – BuildSys 2014
  • 2. [Full Paper] N Sengupta, S Aloka, B Narayanaswamy, H Ismail, S Mathew, " Time Series Data Mining for demand side Decision Support” Innovative Smart Grid Technologies- Asia (ISGT Asia), 2013
  • 3. [Full Paper] DP Seetharam, V Arya, D Chakraborty, ... "Cyber Physical Systems for Smarter Energy Grids: Experiences at IBM Research—India". Journal of the Indian Institute of Science 2013
  • 4. [Full Paper] N Sengupta, K Das, TS Jayram, DP Seetharam, "Optimal Allocation of Land Area for a Hybrid Solar Wind Power Plant". IEEE SmartGridComm 2012
  • 5. [Full Paper] T Bapat, N Sengupta, S Ghai, V Arya, YB Shrinivasan, DP. Seetharam, "User-sensitive Scheduling of Home Appliances". ACM SIGCOMM Green Networking Workshop 2011
  • 6. [Full Paper] VT Chakaravarthy, S Roy, Y Sabharwal, N Sengupta, "Maximizing Throughput of Jobs with Multiple Resource Requirements". International Conference on High Performance Computing 2011
  • Projects
  • 1. Developed website containing advanced charting and reporting functions with Java backend support. Responsible for usage, maintenance, and administration of Informix time series database with huge volumes of sensor data, software support for multiple hardware sensors and integration of the same into the system
  • 2. Responsible for creation, administration, and maintenance of high performance Informix time series database. Developed Java backend for real time relay of data and demand response
  • 3. Yupik - Responsible for writing optimization modules for user-sensitive scheduling of home appliances

Education

Indian Institute of Technology, Delhi

Doctor of Philosophy (Ph.D.) — Computer Science

Jan 2014Jan 2018

Delhi University

Master of Computer Applications (MCA) — Computer Science

Jan 2008Jan 2011

Sri Venkateswara College, University of Delhi

Bachelor of Science (B.Sc.) — Physics

Jan 2005Jan 2008

The Mother's International School

Jan 1997Jan 2005

Stackforce found 100+ more professionals with Machine Learning & Algorithms

Explore similar profiles based on matching skills and experience