# Himanshu Parashar

Himanshu Parashar is a remote worker on Himalayas, Lead Data Scientist at Cotality India, based in India.

Source: https://himalayas.app/@himanshuparashar1

```text
Lead Data Scientist at Cotality building agentic research workflows and bilingual RAG systems for Hindi and English publications.
```

| Detail | Value |
| --- | --- |
| Availability | Looking for a job |
| Location | India |
| Authorized to work in | India |
| Salary expectations | 4548k+ INR |
| Job categories | Data Scientist, Machine Learning Engineer, Product Manager |
| Skills | Data Science, Machine Learning, Agentic AI, RAG, Generative AI, MLOps, Time Series Analysis, Deep Learning, Predictive Analytics, Collaborative Filtering, Association Mining, Causal Inference, Model Explainability, Transformers, Convolutional Neural Networks (CNN) |

## About

```text
At Cotality, I delivered an agentic AI research platform that takes a problem statement through research, drafting, citation verification, originality screening, and publisher formatting. I also engineered safeguards for confidentiality and research integrity.

I scaled a bilingual RAG workflow to 3,500+ books and 1,500+ magazines in Hindi and English, with automated web scraping, OCR cleaning, and Qdrant indexing. I developed a source-aware QA system using hybrid retrieval and GPT-4.1-nano.

At Mercedes-Benz Research and Development India, I developed a Temporal Fusion Transformer to predict EV battery-cell degradation. The approach reduced testing time from over two years to four months.

I also developed time-series classification and vehicle prediction systems, including a CNN that achieved around 98% accuracy and collaborative filtering with 0.1% MAPE. My research includes published work on EV charging reliability and diagnostic trouble-code analysis.
```

## Experience

### Lead Data Scientist (current)

Cotality India. Feb 2026 - Present

```text
Delivered a multi-agent research platform for publication assistance and engineered enterprise safeguards for confidentiality and research integrity. Scaled a bilingual RAG pipeline to 3,500+ books and 1,500+ magazines, and engineered a GPT-4.1-nano QA system with hybrid retrieval.
```

Skills: Python, Google-ADK, Gemini, fastAPI, SQLModel, React, Lang-Chain, Qdrant, MuRIL, GPT-4.1-Nano, RAG, Agentic-AI

### Senior Data Scientist

Mercedes-Benz Research and Development India. Apr 2023 - Feb 2026

```text
Developed a CNN architecture for time-series classification with 98% accuracy and applied saliency maps for critical temporal regions. Implemented a Temporal Fusion Transformer for EV battery degradation prediction, reducing testing time from over 2 years to 4 months. Designed a user-based collaborative filtering framework for mileage prediction with 0.1% MAPE.
```

Skills: Python, PyTorch, TensorFlow, Keras, CNN, Temporal-Fusion-Transformers, LSTM, Attention, Collaborative-Filtering, Time-Series-Analysis, Machine-Learning

### Data Scientist

Mercedes-Benz Research and Development India. Apr 2021 - Mar 2023

```text
Developed a Bayesian-based Probabilistic Graphical Model to predict vehicle registration dates with 0.079% MAPE. Implemented an association mining framework for Diagnostic Trouble Codes with over 99% confidence, and devised a tool for causal analysis of DTCs using predictive analytics and explainability.
```

Skills: Python, Bayesian-Networks, Probabilistic-Graphical-Models, Association-Mining, Predictive-Analytics, Explainability, Machine-Learning

## Education

### Indian Institute of Technology, Kharagpur

Master of Technology, Industrial Engineering and Management. 2018 - 2020. Grade: 8.30/10

```text
Master of Technology in Industrial Engineering and Management with a CGPA of 8.30/10.
```

## Tech stack

- GitHub
- PySpark
- Databricks
- Azure DevOps
- React
- Python
- Azure Machine Learning
- TensorFlow
- PyTorch
- MLflow
- Keras
- FastAPI
- Grafana
- Transformers
- Gemini
- SQL
- Qdrant
- LangChain
- LangGraph
- SQLModel
- Retrieval-Augmented Generation (RAG)

To contact Himanshu, message them from this profile with a Himalayas employer account: [Hire on Himalayas](https://himalayas.app/hire).

## People also viewed

- [Sayan Chaudhuri](https://himalayas.app/@sayanchaudhuri): Data Scientist, Picarro
- [Pradeep Tidke](https://himalayas.app/@pradeeptidke): Senior Data Scientist, Tata Motors Limited
- [Kunal Bhardwaj](https://himalayas.app/@kunalbhardwaj): Data Scientist, HRS Group
- [Saksham Jain](https://himalayas.app/@sakshamjain7): Lead GenAI Engineer (Founding Team), Dionce AI
- [Ayesha Siddiqha](https://himalayas.app/@ayeshasiddiqha): AI/ML Intern, Rubixe
- [mohan krishna](https://himalayas.app/@mohankrishna8): Data Scientist, Fractal
- [HARSH RAJ](https://himalayas.app/@harshraj13): Data Scientist, Verdantas
- [PRATHAM PAYRA](https://himalayas.app/@prathampayra): Doctoral Researcher, Indian Statistical Institute
- [Devansh Kaushik](https://himalayas.app/@devanshkaushik): Claim Analyzer, Mercedes Benz Research & Development India
- [Deepesh Reejwani](https://himalayas.app/@deepeshreejwani): Data Scientist II, Jaguar Land Rover India
