Aasavari Kakne â AI Researcher
Hello đ⨠I am a Senior AI Research Engineer at Attentive AI where I lead a team of talented AI Research Engineers and Subject Matter Experts to deliver AI-automated steel takeoffs. We recently achieved 51% TAT reduction for our steel clients and are baking many exciting features for year 2026. Stay tuned for more updates!⨠I worked as an AI engineer at Intelâs AISE applied ML team where I focus on LLMs (đđđ -đŽđ-đŽ-đˇđđąđ´đ˛ đłđźđż đźđ˝đ˛đťđđźđđżđ°đ˛ đŽđťđą đŽđťđťđźđđŽđđśđźđť đłđżđ˛đ˛ đ˛đđŽđšđđŽđđśđźđť đźđł đĽđđ on Gaudi2). Prior to this, I worked on 'Preprocessing pipeline of large datasets for LLM pretraining comprising of Near deduplication, PII removal and Global deduplication'. We quantified efficacy of our Near Deduplication and benchmarked it at F-1 score of 0.99861. ⨠I also dabble in GNNs and am grateful to receive âđđđŚđ đđ˛đ˝đŽđżđđşđ˛đťđ đĽđ˛đ°đźđ´đťđśđđśđźđť đđđŽđżđąâ for 2024 for being a core contributor to our RecSys 2023 submission that ranked #2 đśn the industry leaderboard. đ˘đđż đđźđšđđđśđźđť đśđ đ˝đđŻđšđśđđľđ˛đą đśđť đđđ - âGraph Enhanced Feature Engineering for Privacy Preserving Recommendation Systemsâ . đŞđ˛ đŽđšđđź đłđśđšđ˛đą đŽ đ¨đŚ đ˝đŽđđ˛đťđ for âConstruction of bipartite graphs from Privacy Preserved datasetsâ as part of our RecSys 2023 submission. âď¸ Prior to Intel, I completed đ đŚ đŽđ đŚđđŽđťđłđźđżđą in Applied Math. I appreciated learning from classes like - đĄđđŁ, đĄđđ¨, đđŠ, đđĄđĄđ đŽđťđą đ đ. Our blog âFantastic graphs and how to complete themâ for GNN class was selected to be published by Prof. Jure Leskovec at Stanford. đ đŽđšđđź đ˛đťđˇđźđđ˛đą đ§đđśđťđ´ đłđźđż đ°đšđŽđđđ˛đ - đŚđźđłđđđŽđżđ˛ đđťđ´đśđťđ˛đ˛đżđśđťđ´, đđąđđŽđťđ°đ˛đą đŚđźđłđđđŽđżđ˛ đđťđ´đśđťđ˛đ˛đżđśđťđ´ đŽđťđą đĄđđ¨ đŽđ đŚđđŽđťđłđźđżđą. In this time, I also got to work on - Weak Supervised networks, Semi Supervised networks, Multimodal representation learning, Transfer Learning techniques, Generative Adversarial Networks, and Knowledge Graph completion during my projects and internships. âď¸ Prior to Stanford, I completed đ.đ§đ˛đ°đľ đśđť đ đŽđđľ đŽđťđą đđźđşđ˝đđđśđťđ´ đŽđ đđđ§ đđ˛đšđľđś where my favourite classes were probability, statistics and linear algebra. đ đšđźđđ˛đą đđź đ§đ đłđźđż - đđđŽđđśđđđśđ°đŽđš đśđťđłđ˛đżđ˛đťđ°đ˛ đŽđťđą đ°đŽđšđ°đđšđđ đ°đšđŽđđđ˛đ đŽđ đđđ§-đ. During my third year at IIT-D, I studied at University of Waterloo as part of IIT-Dâs foreign exchange program. UWâs Differential equations and âlearning how to learnâ classes were amazing.
Stackforce AI infers this person is a Senior AI Research Engineer specializing in machine learning and data analysis for AI-driven solutions.
Location: Noida, Uttar Pradesh, India
Experience: 4 yrs
Skills
- Machine Learning
- Deep Learning
- Data Analysis
Career Highlights
- Achieved 51% TAT reduction for steel clients.
- Ranked #2 in industry leaderboard for RecSys 2023.
- Published research on privacy-preserving recommendation systems.
Work Experience
Attentive.ai
Senior AI Research Engineer (5 mos)
Research Engineer II (11 mos)
Intel Corporation
AI software solutions engineer (2 yrs 3 mos)
Stanford University
Teaching Assistant, Natural Language Understanding (6 mos)
Teaching Assistant, Advanced Software Development (2 mos)
Teaching Assistant, Software Development (3 mos)
Sprinklr
Machine Learning Intern (5 mos)
NPTEL
Teaching Assistant (1 mo)
Indian Institute of Technology, Delhi
Teaching Assistant, freshmen Calculus (4 mos)
NoBroker.com
Automatic Description Generation | Data Science Intern (2 mos)
Education
Master of Science - MS at Stanford University
Bachelor of Technology - BTech at Indian Institute of Technology, Delhi
Exchange student at University of Waterloo
