Nikhil Kulkarni

AI Researcher

Mountain View, California, United States10 yrs 9 mos experience

Highly StableAI Enabled

Key Highlights

Key contributor to AI-powered search at Target.
Presented Generative AI insights to senior leadership.
Developed scalable AI agents at Google.

Stackforce AI infers this person is a SaaS and E-commerce AI specialist with expertise in Natural Language Processing and Machine Learning.

Contact

Skills

Core Skills

Artificial Intelligence (ai)Information RetrievalNatural Language ProcessingDeep LearningMachine Learning

Other Skills

AI AgentsAI powered searchAlgorithmsApache KafkaApache SparkBig DataCC++Data MiningData ScienceData StructuresData VisualizationDigital Image ProcessingDigital PhotographyDigital Signal Processing

About

Currently Applied AI Engineer at Google working on Gemini for Enterprise - AI-Powered enterprise Search, Assistant, AI Agents experience powered by Gemini models. Previously, Lead AI Scientist At Target’s (Bay area) Search Team. Built Query Understanding and Ranking models for E-commerce Search. Utilized Generative AI, Large language models, Graph neural networks, RAG, Natural Language Processing with Deep Learning, Ranking, Evaluation techniques. The Johns Hopkins University Alum with a focus on Machine Learning, Information Retrieval, and Natural Language Processing. More: https://nikhilk.me/

Experience

Google

Applied AI Engineer

Oct 2024 – Present · 1 yr 5 mos · Sunnyvale, California, United States · On-site

Gemini for Enterprise and AI Agents lead on a product-driven Applied AI Engineering team within Google Cloud AI. We build Search, Generative AI Assistant, Agents for enterprises powered by Gemini. Bridging the last-mile gap between the latest research developments and commercial deployments. We use end-to-end system design to develop vertically-integrated, scalable, industry-specific and differentiated agents for Google and Google Cloud customers. My work is at the intersection of Gemini for Enterprise Search + Assistant Quality, ADK agents, Agent Engine.
Gemini for Enterprise is the launch point for enterprise-ready AI agents, helping increase employee productivity for complex tasks through google quality search, knowledge graph and gemini's advanced reasoning with one single prompt.

AI AgentsAI powered searchInformation RetrievalArtificial Intelligence (AI)Enterprise Search

Target

2 roles

Lead AI Scientist - Search

Promoted

Aug 2020 – Sep 2024 · 4 yrs 1 mo

Related job titles: Machine Learning Engineer, Machine Learning Scientist, Applied Scientist, Research Engineer, Deep Learning Engineer, Search Relevancy Engineer, NLP Engineer, Software Engineer - Machine Learning
Spent 6+ years in the AI Search team within the Data Sciences Org. 100% of the search queries on target.com pass through my team's ML models.
Built Query Understanding, Ranking models using LLMs, Deep Learning techniques to improve the e-commerce search experience for millions of Target users. Part of the team that delivered Search 3.0 (AI powered search) in 2024 and Search 2.0 (Deep learning based search) in 2018.
Designed, developed and deployed Query Intent + Attribute Classifiers trained using deep custom built CNNs
Query Equivalence, Query Rewrite, Query Expansion, Automatic synonym detection trained using Query Reformulation Graph + GNN link prediction models
Query Expansion, Query Router trained using LLMs
Responsibilities:
Identifying search related business problems, collecting / generating training data, training deep learning models, writing ML pipelines, evaluating models, serving via REST APIs, integrating in the search system, designing & analyzing A/B tests, presenting results to leadership
Achievements:
2024 Key contributor to the Natural Language Search project that leverage GenAI tech to improve search experience of abstract search queries like - sofa, summer, candy, gym, shark birthday
2023 Presented Demystifying Generative AI to 641 technical colleagues, including SVPs and VPs of different verticals.
2021 Increased search attributable demand by $$$ Millions through multiple Query Understanding models
2019 Received Chief Data Analytics Officer Award for enabling e-commerce search with data science. Increased search attributable demand by 4.8%, Relevancy by 2% YoY.
Areas:
Information Retrieval, Search, Natural Language Processing, Deep Learning, Machine Learning, Graph Neural Network, Large Language Models, GenAI

Natural Language ProcessingDeep LearningGenerative AIInformation RetrievalAlgorithmsTechnical Leadership+5

Senior AI Scientist - Search

Mar 2018 – Aug 2020 · 2 yrs 5 mos

Search Query Understanding and Ranking Team
Search, Ranking, Evaluation, NLP, Deep learning, Information Retrieval

Github

Machine Learning Intern

Jun 2017 – Sep 2017 · 3 mos · San Francisco Bay Area

https://internships.github.com/interns/2017/nikhil
Used Machine Learning and Information Retrieval based techniques to recommend similar discussion threads on GitHub. Worked on 84 Million public issues. Designed a model for recommending semantically similar issues. The model solves the search and recommendation problem for GitHub Issues. It will serve as a baseline model for the Machine Learning team going forward.

Troubleshooting

Johns hopkins university

Graduate Course Assistant

Sep 2016 – Dec 2016 · 3 mos · Baltimore, Maryland

Project Advisor and Course Assistant of EN.600.421 Object Oriented Software Engineering course.

Innoplexus

Data Scientist

Mar 2016 – Aug 2016 · 5 mos · Pune, Maharashtra, India

Worked mainly on Information Retrieval, NLP, Machine Learning. Designed and trained a custom Named Entity Recognition system to tag biomedical entities in medical journal abstracts.
Used Conditional Random Fields to tag unknown data set cutting down manual tagging.

Tinyowl

Data Scientist

Apr 2015 – Mar 2016 · 11 mos · Mumbai, Maharashtra, India

Worked on a project involving food taxonomy generation and text mining. Designed a pipeline to categorize food item names into different categories such as cuisine type, meal type, veg/non-veg etc. Achieved 93% avg. precision for the taxonomy based on 7 different categories. Received special recognition from management.
(Acquired by Zomato)

Zlemma

2 roles

Data Scientist

Jun 2014 – May 2015 · 11 mos · Pune India / Palo alto US

Worked on Resume Parser project which involved pattern recognition and machine learning techniques. Improved patent-pending algorithms by tuning parameters using statistical data analysis. Analyzed valuation curves using cases where the algorithm worked best and worst. Problem statements involved an application of Machine Learning in Natural Language Processing.
(Acquired by Hired.com)