Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I am an AI/ML engineer with hands-on experience across the end-to-end ML lifecycle, from data preprocessing and feature engineering to model training, evaluation, API integration, and production deployment. I’m proficient in Python, Scikit-learn, PyTorch, TensorFlow, FastAPI, and React, with strong interests in computer vision, NLP, and generative AI. I enjoy exploring LangChain, RAG, vector databases, and agentic AI to build scalable, data-driven solutions. I’m passionate about translating complex problems into robust, user-friendly AI systems. I continuously expand my toolkit with modern ML frameworks, API-driven architectures, and explainability methods (SHAP, LIME), while keeping a focus on measurable impact and stakeholder communication.…I am an AI/ML engineer with hands-on experience across the end-to-end ML lifecycle, from data preprocessing and feature engineering to model training, evaluation, API integration, and production deployment. I’m proficient in Python, Scikit-learn, PyTorch, TensorFlow, FastAPI, and React, with strong interests in computer vision, NLP, and generative AI. I enjoy exploring LangChain, RAG, vector databases, and agentic AI to build scalable, data-driven solutions. I’m passionate about translating complex problems into robust, user-friendly AI systems. I continuously expand my toolkit with modern ML frameworks, API-driven architectures, and explainability methods (SHAP, LIME), while keeping a focus on measurable impact and stakeholder communication.

Mansha Kshatriya

Data Scientist, AI Engineer, AI Developer, +2





I am an AI/ML engineer with hands-on experience across the end-to-end ML lifecycle, from data preprocessing and feature engineering to model training, evaluation, API integration, and production deployment. I’m proficient in Python, Scikit-learn, PyTorch, TensorFlow, FastAPI, and React, with strong interests in computer vision, NLP, and generative AI. I enjoy exploring LangChain, RAG, vector databases, and agentic AI to build scalable, data-driven solutions. I’m passionate about translating complex problems into robust, user-friendly AI systems. I continuously expand my toolkit with modern ML frameworks, API-driven architectures, and explainability methods (SHAP, LIME), while keeping a focus on measurable impact and stakeholder communication.…I am an AI/ML engineer with hands-on experience across the end-to-end ML lifecycle, from data preprocessing and feature engineering to model training, evaluation, API integration, and production deployment. I’m proficient in Python, Scikit-learn, PyTorch, TensorFlow, FastAPI, and React, with strong interests in computer vision, NLP, and generative AI. I enjoy exploring LangChain, RAG, vector databases, and agentic AI to build scalable, data-driven solutions. I’m passionate about translating complex problems into robust, user-friendly AI systems. I continuously expand my toolkit with modern ML frameworks, API-driven architectures, and explainability methods (SHAP, LIME), while keeping a focus on measurable impact and stakeholder communication.

Available to hire

I am an AI/ML engineer with hands-on experience across the end-to-end ML lifecycle, from data preprocessing and feature engineering to model training, evaluation, API integration, and production deployment. I’m proficient in Python, Scikit-learn, PyTorch, TensorFlow, FastAPI, and React, with strong interests in computer vision, NLP, and generative AI. I enjoy exploring LangChain, RAG, vector databases, and agentic AI to build scalable, data-driven solutions.

I’m passionate about translating complex problems into robust, user-friendly AI systems. I continuously expand my toolkit with modern ML frameworks, API-driven architectures, and explainability methods (SHAP, LIME), while keeping a focus on measurable impact and stakeholder communication.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Language

English

Advanced

Work Experience

AI Engineer / Full-Stack Developer at Verdix

December 1, 2025 - Present

Architected a production-grade computer vision platform end-to-end (async image recognition and classification pipeline using Google Gemini Vision API), automated scoring, and data quality checks. Built a React/TypeScript frontend and Node.js/Express backend, processing 500+ daily analyses with real-time throughput. Implemented structured output validation and accelerated async data pipeline performance, reducing p95 API latency by 40% and improving reliability under production load. Led delivery from requirements to live deployment as a solo engineer, delivering explainable outputs for non-technical stakeholders.

ML Engineer / Geo-Matching Specialist at RaktSetu

August 1, 2025 - Present

Executed end-to-end ML lifecycle for real-time geo-matching: raw data ingestion, preprocessing, feature engineering, classifier training with rigorous evaluation (cross-validation, precision, recall, F1, ROC-AUC), FastAPI deployment, and a JS frontend. Reduced donor–patient match time by ~60% and diagnosed production inference bottlenecks to bring p95 latency under 500ms. Developed a real-time alert pipeline translating a healthcare problem into a scalable, production-ready ML solution.

Clinical ML Data Scientist at PredictoCare

February 1, 2024 - Present

Engineered a 30+ feature preprocessing pipeline on the Wisconsin Breast Cancer dataset; trained and tuned an SVM via GridSearchCV with full model evaluation (cross-validation, precision/recall/F1/ROC-AUC), achieving 97% classification accuracy and reducing false negatives. Integrated SHAP + LIME explainability delivering feature-level reasoning per prediction, enabling interpretation in regulated AI environments.

LLM-NLP API Developer at LinkedIn Post Generator (Personal Project)

February 1, 2024 - Present

Designed and launched an LLM-powered RESTful API (Groq + Llama 3) with NLTK keyword extraction, prompt engineering, structured outputs, and a user-feedback evaluation loop. Achieved a 70% reduction in content creation time per user per week; managed end-to-end pipeline from data ingestion to a Streamlit dashboard deployment.