Hardik Srivastava

Applied Research Scientist

Researching and building large-scale intelligent language systems

hardiksrwork@gmail.com

Connect via Topmate

LeetCode

About

I'm a ML researcher and engineer pursuing an MS in Data Science at the University of Washington, where I focus on building controllable and efficient language-model systems. My interests span LLMs, reinforcement learning, representation learning, and scalable machine learning, with a focus on making intelligent systems more reliable, secure, and efficient.

At the Paul G. Allen School of CSE, I work with Prof. Zaid Harchaoui on controllable language generation, authorship privacy, and autonomous machine learning research. My research addresses the challenge of reducing author-identifying stylistic signals in written text while preserving its original meaning and communicative intent. To tackle this, I am exploring LLM alignment and post-training through feedback-driven evaluation and inference-time control mechanisms to steer model behavior across competing objectives. More recently, I'm looking into building autonomous research agents using reinforcement learning and bandit algorithms with delayed feedbacks to automate ML experimentation and accelerate optimization cycles by 15x.

Prior to UW, I spent over two years as an Applied Scientist at JPMorgan Chase, building and deploying production ML/AI systems for AI Trust & Safety. I've worked across graph learning, multimodal NLP, large-scale entity resolution, fraud detection and distributed ML infrastructure. My work contributed to over $220M in annual fraud savings as well. In summer 2026, I returned to JPMorganChase as an Applied Scientist Intern to work on temporal graph learning for adversarial actor modeling and scalable multi-GPU training.

Earlier, I worked as a MLE at Samsung Research, contributing to production-scale NLU intent routing for the Samsung Bixby assistant and optimizing word-sense disambiguation models through INT8 Quantization-Aware Training. I also conducted research in Human-Computer Interaction and NLP at the Accessible Computing Lab at McGill University as a Mitacs Research Scholar.

I enjoy working at the intersection of research and engineering, taking challenging ML problems from formulation and experimentation to efficient, scalable systems. Open to connect if your team is building something at the intersection of LLM learning, reasoning, and trustworthy AI

Research Interests

My core research interests lie at the intersection of Natural Language Processing, Generative AI, Graph Representation Learning, Human-AI Interaction, Causality, and Trustworthy ML. I focus on designing language and multimodal systems that are interactive, adaptable and explainable to diverse user needs, and capable of operating reliably in high-stakes environments.

Skills

Machine Learning & AI

Probabilistic Reasoning
Generative AI
Graph Learning
Zero & Few-shot Learning
AI Alignment & Interpretability
Contrastive Learning
Ranking & Recommendation

Natural Language Processing

Large Language Models (LLMs)
Transformer Architectures
Retrieval-Augmented Generation (RAG)
Semantic Search & Retrieval
Multimodal Learning

Research Tooling & MLOps

PyTorch
TensorFlow
HuggingFace
MLflow
Scikit-Learn
Weights & Biases
LaTeX

Core Programming & CS

C++
Python
Java
SQL
Data Structures & Algorithms
Distributed Systems

Big Data

Apache Spark
Kafka
Hadoop
ZooKeeper
Splunk

ML Systems & Infra

vLLM
CUDA
Distributed Training
AWS
Docker & Kubernetes
MLflow
CI/CD

News

Publications

Architecture Diagram

Multi-Modal Sentiment Analysis Using Text and Audio for Customer Support Centers

Hardik Srivastava, Sneha Sunil, K. Shantha Kumari, P. Kanmani

ICACTCE: International Conference on Advances in Communication Technology and Computer Engineering (Springer Nature) (2023)

Proposed CM-BERT, a novel multimodal sentiment analysis model that fuses textual and audio features to robustly capture customer sentiment

Architecture Diagram

Decision Support Complaint Prioritization System using a Statistical Multi-Method Algorithmic approach

Hardik Srivastava, Mayank Jha, T. Karthick

Undergraduate Thesis (2023)

Proposed a statistical multi-method framework combining rule-based scoring and algorithmic ranking to integrate diverse decision signals into a coherent prioritization score for efficient re-ranking.

Architecture Diagram

Neural Text Style Transfer with Custom Language Styles for Personalized Communication Systems

Hardik Srivastava, Sneha Sunil, K. Shantha Kumari

ICKECS'22: International Conference on Knowledge Engineering and Communication Systems (IEEE) (2022)

Introduced StyleLM, a style-conditioned neural text transfer model enabling fine-grained control over language style for personalized communication.

Architecture Diagram

Automatic Screening and Staging of Multi-Stage Diabetic Retinopathy using Deep Learning techniques

Hardik Srivastava, T. Rajalakshmi

Preprint (2022)

Proposed a deep learning framework for automated detection and staging of multi-stage diabetic retinopathy.

Architecture Diagram

Using NLP Techniques for Enhancing Augmentative and Alternative Communication Applications

Hardik Srivastava

IJREAM: International Journal for Research in Engineering Application & Management (UGC) (2021)

NLP-based framework to enhance AAC applications, enabling dynamic vocabulary expansion and more expressive sentence generation.

Architecture Diagram

Multi-Modal Sentiment Analysis Using Text and Audio for Customer Support Centers

Hardik Srivastava, Sneha Sunil, K. Shantha Kumari, P. Kanmani

ICACTCE: International Conference on Advances in Communication Technology and Computer Engineering (Springer Nature) (2023)

Proposed CM-BERT, a novel multimodal sentiment analysis model that fuses textual and audio features to robustly capture customer sentiment

Architecture Diagram

Neural Text Style Transfer with Custom Language Styles for Personalized Communication Systems

Hardik Srivastava, Sneha Sunil, K. Shantha Kumari

ICKECS'22: International Conference on Knowledge Engineering and Communication Systems (IEEE) (2022)

Introduced StyleLM, a style-conditioned neural text transfer model enabling fine-grained control over language style for personalized communication.

Architecture Diagram

Using NLP Techniques for Enhancing Augmentative and Alternative Communication Applications

Hardik Srivastava

IJREAM: International Journal for Research in Engineering Application & Management (UGC) (2021)

NLP-based framework to enhance AAC applications, enabling dynamic vocabulary expansion and more expressive sentence generation.

Architecture Diagram

Decision Support Complaint Prioritization System using a Statistical Multi-Method Algorithmic approach

Hardik Srivastava, Mayank Jha, T. Karthick

Undergraduate Thesis (2023)

Proposed a statistical multi-method framework combining rule-based scoring and algorithmic ranking to integrate diverse decision signals into a coherent prioritization score for efficient re-ranking.

Architecture Diagram

Automatic Screening and Staging of Multi-Stage Diabetic Retinopathy using Deep Learning techniques

Hardik Srivastava, T. Rajalakshmi

Preprint (2022)

Proposed a deep learning framework for automated detection and staging of multi-stage diabetic retinopathy.

Architecture Diagram

Decision Support Complaint Prioritization System using a Statistical Multi-Method Algorithmic approach

Hardik Srivastava, Mayank Jha, T. Karthick

Undergraduate Thesis (2023)

Proposed a statistical multi-method framework combining rule-based scoring and algorithmic ranking to integrate diverse decision signals into a coherent prioritization score for efficient re-ranking.

Architecture Diagram

Multi-Modal Sentiment Analysis Using Text and Audio for Customer Support Centers

Hardik Srivastava, Sneha Sunil, K. Shantha Kumari, P. Kanmani

ICACTCE: International Conference on Advances in Communication Technology and Computer Engineering (Springer Nature) (2023)

Proposed CM-BERT, a novel multimodal sentiment analysis model that fuses textual and audio features to robustly capture customer sentiment

Architecture Diagram

Automatic Screening and Staging of Multi-Stage Diabetic Retinopathy using Deep Learning techniques

Hardik Srivastava, T. Rajalakshmi

Preprint (2022)

Proposed a deep learning framework for automated detection and staging of multi-stage diabetic retinopathy.

Architecture Diagram

Neural Text Style Transfer with Custom Language Styles for Personalized Communication Systems

Hardik Srivastava, Sneha Sunil, K. Shantha Kumari

ICKECS'22: International Conference on Knowledge Engineering and Communication Systems (IEEE) (2022)

Introduced StyleLM, a style-conditioned neural text transfer model enabling fine-grained control over language style for personalized communication.

Architecture Diagram

Using NLP Techniques for Enhancing Augmentative and Alternative Communication Applications

Hardik Srivastava

IJREAM: International Journal for Research in Engineering Application & Management (UGC) (2021)

NLP-based framework to enhance AAC applications, enabling dynamic vocabulary expansion and more expressive sentence generation.

Projects

Recursive Episodic Memory (REM) RAG
Engineered a memory compression framework for long-context LLM agents. Utilizing FAISS-based dense retrieval, hierarchical semantic indexing, and adaptive replay mechanisms, this architecture mitigates context decay and enables persistent, multi-session agentic reasoning across complex computational pipelines.
PokerFace: RL for Strategic Deception in LLMs
Trained LLMs for strategic deception and negotiation in a Market for Lemons env using PPO. Engineered a two-stage reasoning pipeline to maximize strategic rewards without leaking hidden states to opponents. Scaled training via QLoRA for parameter-efficient fine-tuning on instruct models.
Ad-Astra: Agentic Ad Intelligence Engine
Built a dynamic ad generation and placement engine for X (Twitter), using Grok 4.1 Reasoning models. Utilized behavioral scroll telemetry and temporal interaction signals to optimize in-feed ad placement improving relevance and engagement.
KeyCognition: Predicting Cognitive Load via Keystrokes
Behavioral AI platform that predicts user cognitive load in real-time using temporal sequence modeling. Engineered a React app to extract micro-temporal typing anomalies like dwell time & flight time using 136M keystrokes.
Auto-Explainability & Evaluation of LLMs
Transformer explainability framework designed to demystify language model generation. Utilizes Integrated Gradients to map and visualize token-level prediction attributions enabling deep interpretability.
No-Code Algo-Trading Portfolio Manager
Domain-Specific Language (DSL) framework powering a no-code trading platform. Leverages hybrid recommendation models to dynamically compile drag-and-drop user inputs into optimized algorithms, thereby minimizing risk and expected shortfall.

Articles

MLOps 101 — Machine Learning Workflow Orchestration using TensorFlow Extended
Deploying Machine Learning workloads efficiently in Production using TFX
A/B Testing in Data Science
Learning how to let the customer do the talking, without actually talking.

Vitæ

Here is my Resume.

  • JPMorganChase June - Sept 2026
    Applied Scientist Intern
    Temporal Graph-ML for Adversarial Actor Modeling
  • University of Washington Sept 2025 - Present
    Graduate Research Assistant
    Optimizing LLM Alignment and Interpretability
  • University of Washington Sept 2025 - Present
    MS Student
    Masters in Data Science
  • JPMorganChase May - Aug 2025
    Applied Scientist II
    Fraud Prevention, Graph ML and Contrastive Pre-Training
  • JPMorganChase June 2023 - May 2025
    Applied Scientist
    Fraud Intelligence and Representaion Learning
  • JPMorganChase. Feb - June 2023
    Software Engineer Intern - AI/ML
    Elastic Search and Semi-Supervised Ranking
  • McGill University June - Sept 2022
    Research Intern
    NLP/HCI at Accessible Computing Lab
  • Samsung Global Research Jan - June 2022
    Machine Learning Engineer - Intern
    Samsung Bixby's search optimization using Quantization-Aware Training
  • Inventuers Sept - Nov 2023
    Founder & Tech Lead
    Co-founded and released Curae, a platform to streamline healthcare services
  • DRDO, Govt. of India Aug - Nov 2021
    Research Intern
    Video Stabilization and Denoising for Long-Range Camera-feeds in Military Drones
  • SoftSensor.ai Apr - Oct 2021
    Data Scientist Intern
    Pattern Recognition in Whole Slide Imaging (Lymphoma)
  • SRM University Jan - July 2021
    Undergraduate Research Assistant
    Predictive Modeling for Diabetic Retinopathy detection
  • SRM University May 2019 - June 2023
    B.Tech Student
    Computer Science Engineering with a specialization in Big Data Analytics