Sachin Rathod

Data Scientist

Bengaluru, Karnataka, India9 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Developed scalable AI platforms for automotive pricing.
Achieved 94% accuracy in vehicle damage detection.
Created innovative Generative AI solutions.

Stackforce AI infers this person is a Data Scientist specializing in Automotive AI solutions and MLOps.

Contact

Skills

Core Skills

MlopsComputer VisionGenerative AiNatural Language Processing (nlp)

Other Skills

Data ScienceCloud ComputingMachine LearningDeep LearningImage AnalysisImage ProcessingFlaskNLPOCRGoogle Cloud Platform (GCP)Retrieval-Augmented Generation (RAG)GitMachine VisionData AnalysisMicrosoft Azure

About

I am a Data Scientist with strong expertise in data preprocessing, MLOps pipeline development, and building end-to-end AI/ML solutions. At Mahindra AI, I collaborate with a team partnering with Mahindra First Choice Wheels Ltd. (MFCWL) to develop innovative and scalable AI platforms, with hands-on experience in the second-hand car pricing domain. Starting my career as a Data Science Intern, I have progressed to my current role through consistent performance and learning. My core strengths include designing microservices, optimizing MLOps pipelines, managing cloud infrastructure on AWS and GCP, and deploying production-grade AI models. I bring deep technical proficiency in Computer Vision, Natural Language Processing (NLP), Generative AI (GenAI), Python, and C++. I hold a B.Tech in Artificial Intelligence and Data Science from Dr. D. Y. Patil University, graduating with honors.

Experience

9 mos

Total Experience

9 mos

Average Tenure

9 mos

Current Experience

Mahindra.ai

2 roles

Data Scientist

Aug 2025 – Present · 9 mos · On-site

Pricing Engine · Live in Production
Built a scalable microservice-based pricing engine with a hybrid ML + statistical model, using market and internal data for accurate pricing and LTV decisions
Served 45+ clients (banks & OEMs), handling 900K+ monthly requests with high reliability
Vehicle Damage Detection System · Live in Production
Designed and trained multiple deep learning CV models end-to-end — part segmentation, background masking, and damage detection — processing 360° vehicle images to auto-detect, localize, and generate a structured part-wise damage map
Eliminated manual inspection dependency and standardized the process with 94% model accuracy
Scaled to 3 lakh+ inspections/month with response time under 3 seconds per request
AutoVision Studio
Created a Generative AI solution to transform product images into high-quality while maintaining originality , studio-style visuals using intelligent masking and prompt-based enhancements.

Generative AIData ScienceCloud ComputingMLOpsComputer VisionMachine Learning

Data Science Intern

May 2024 – Jul 2025 · 1 yr 2 mos · On-site

Entity Mapping for Pricing Engine · Live in Production
Built and deployed an end-to-end entity resolution system using a custom embedding model and hybrid retrieval (semantic + phonetic + fuzzy matching)
Achieved 98% accuracy on noisy, unstructured data with zero standardization
AI-Powered VIN Fraud Detection · Live in Production
Developed a GenAI-based image analysis system (OCR + tamper/anomaly detection) to detect fraud in VIN/chassis plates
Delivered 95% precision, replacing a 20–25 member manual review process
Prevents high-value fraud (₹40–45L exposure per case) with explainable, standardized decisions

Natural Language Processing (NLP)FlaskData Science