Portfolio

Pramod Chavhan

Pramod.

Resume Let's talk

Open to full-time · Remote & hybrid · Pune

Pune, Maharashtra, India · 3.6 yrs · Data Scientist | GenAI & ML Engineer

PramodChavhan

Production RAG chatbots & ML systems — shipped at Rutamsoft and Pipran Infotech with measurable business impact.

View selected work

Download CV

Contact

Experience

3.6 yrs

Featured work

6 projects

Certifications

16 earned

Current role

Rutamsoft Technology LLP

Selected work

Projects that shipped

Case studies from production — RAG chatbots, computer vision, IoT, and predictive ML with measurable impact.

Featured · GenAIEnterprise · NDA

Rutamsoft Technology LLP · Nov 2025 – Present

QuotePlan EVA AI Chatbot

FastAPI RAG chatbot for QuotePlan users — document Q&A, project lookup, pending PO tracking, P&L analysis, and project creation via natural language with permission-based actions.

Problem

QuotePlan users spent too long hunting documents, POs, and project data across a complex ERP-style product.

Approach

Built a FastAPI RAG stack with Qdrant vector search, sentence-transformer embeddings, and permission-aware tool calls so users can query and act in natural language.

→Document Q&A, project lookup, and pending PO tracking in one chat interface
→P&L analysis and project creation via NL commands with role-based access
→Live in production at Rutamsoft — enterprise deployment under NDA

↓ Manual

Search Effort

RAG + Qdrant

Stack

Live

Status

PythonFastAPIOpenAI APIQdrantSentence TransformersJavaScript

Production · available on request

Featured · RAGEnterprise · NDA

Pipran Infotech Pvt Ltd · Aug 2025 – Sep 2025

PDF Semantic Search (RAG)

Retrieval-Augmented Generation pipeline for large PDFs — PyPDF2 ingestion, OpenAI embeddings, Pinecone similarity search, and FastAPI answers with 95% relevancy on 500+ page docs.

Problem

Teams couldn't quickly find answers inside 500+ page PDF manuals — keyword search missed context and wasted hours.

Approach

Designed a RAG pipeline: PyPDF2 chunking, OpenAI embeddings, Pinecone similarity search, and FastAPI endpoints for semantic Q&A.

→95% answer relevancy on large technical documents
→Search time reduced by ~70% vs manual lookup
→Scalable ingestion pipeline for new document uploads

-70%

Search Time

95%

Relevancy

500+ pp

Doc Size

PythonOpenAI APIPineconePyPDF2FastAPINLP

Production · available on request

Featured · ML

Pipran Infotech Pvt Ltd · Apr 2023 – Sep 2023

ML Appointment Scheduler

Flask + React home sample collection scheduler for diagnostic centers — Google Maps agent assignment, Random Forest delivery prediction, Twilio SMS, and Firebase booking. Cut wait times by 20% with 90% on-time delivery.

Problem

Diagnostic centers struggled with manual scheduling, long patient wait times, and unpredictable phlebotomist routing.

Approach

Shipped Flask + React app with Google Maps agent assignment, Random Forest delivery prediction, Twilio SMS alerts, and Firebase booking sync.

→20% reduction in patient wait times
→90% on-time sample collection rate
→50+ daily home collections handled reliably

-20%

Wait Times

Daily Collections

90%

On-time

PythonScikit-learnPandasFlaskReactGoogle Maps API

Live demo Code

Featured · CV

Pipran Infotech Pvt Ltd · Oct 2023 – Mar 2024

Face Recognition Attendance

Real-time attendance with OpenCV and insightFace ResNet-34 — Flask backend, Firebase logging, and React admin dashboard for 1000+ students. Reduced manual effort by 60% and improved accuracy by 40%.

Problem

Manual college attendance was slow, error-prone, and couldn't scale beyond 1000+ students.

Approach

Built real-time face recognition with insightFace ResNet-34, OpenCV pipeline, Flask API, Firebase logging, and React admin dashboard.

→60% less manual admin effort
→40% accuracy improvement over legacy process
→1000+ students tracked in production

-60%

Manual Effort

+40%

Accuracy Lift

1000+

Students

PythonOpenCVinsightFace ResNet-34FlaskFirebaseReact

Live demo Code

Featured · IoT

Pipran Infotech Pvt Ltd · Apr 2024 – Jan 2025

Smart Irrigation System

IoT irrigation for Punjab farms — Raspberry Pi sensors, Random Forest predictions via OpenWeather API, and Flask remote access. Increased crop yield by 15% and reduced water usage by 25% across 10 hectares.

+15%

Crop Yield

-25%

Water Saved

10 ha

Coverage

PythonIoT SensorsScikit-learnFlaskRaspberry PiOpenWeather API

Live demo Code

Featured · NLP

Pipran Infotech Pvt Ltd · Feb 2025 – Jul 2025

BERT Question Answering

Fine-tuned BERT for medical document Q&A — PDFs, Word, JSON via FastAPI, with Pinecone storage and sub-2s answers for clinicians. Improved query accuracy by 40%.

+40%

Query Accuracy

<2s

Latency

BERT + Pinecone

Stack

PythonFastAPIBERTTransformersPineconeNLP

Live demo Code

More experiments

Second Hand Car Price Prediction

ML model predicting used-car market value from make, model, year, mileage, and condition for buyers and sellers.

PythonScikit-learnXGBoost

Forest Fire Prediction

Regression model predicting forest fire spread and intensity from meteorological and geographical features.

PythonRandom ForestGradient Boosting

IoT

Credit Score Prediction

Classification model for creditworthiness from financial history, payment patterns, and demographics.

PythonScikit-learnLightGBM

Backorder Prediction

Supply-chain classifier identifying products at risk of backorder from inventory, lead time, and demand signals.

PythonScikit-learnCatBoost

House Price Prediction

Residential price regression from location, size, amenities, and market trends for data-driven valuations.

PythonScikit-learnXGBoost

Breast Cancer Survival Prediction

Survival analysis model using clinical and demographic features to support oncology prognoses.

PythonScikit-learnSurvival Analysis

About

Building AI that ships

Results-oriented Data Scientist with 3.6 years of experience delivering production-ready ML and GenAI solutions. Proven track record in building and deploying machine learning models, LLM-based applications (RAG, fine-tuning, Hugging Face), and end-to-end data pipelines using Python, Flask, and cloud technologies (Azure, AWS, Vector Database). Adept at solving real-world business problems using AI with measurable impact — including automation, efficiency improvement, and data-driven decision-making.

Oct 2025 – Current

Rutamsoft Technology LLP

AI/ML Developer

Mar 2023 – Sep 2025

Pipran Infotech Pvt Ltd

Data Science Associate

Apr 2022 – Sep 2022

Innodatastics

Data Science Intern

Master of Computer Applications (MCA)

2022 – 2024 · 8.2 CGPA

Trinity Academy of Engineering, Pune

Bachelor of Computer Applications (BCA)

2019 – 2021 · 7.8 CGPA

Gopikabai Sitaram Gawande College, Umarkhed

Languages: English · Marathi · Hindi

Download CV

Career

Experience

Oct 2025 – Current

Rutamsoft Technology LLP

AI/ML Developer

Pune, India

Developed a RAG-based AI chatbot for QuotePlan to manage product queries, workflows, and automation.
Enabled end-to-end project handling via chatbot, including process guidance and system information.
Automated vendor pricing, stock queries, and CRUD operations using natural language commands.
Integrated chatbot with backend systems for real-time and secure data operations.

Mar 2023 – Sep 2025

Pipran Infotech Pvt Ltd

Data Science Associate

Pune, India

Designed & deployed 7+ ML models (classification, regression) using Python, Scikit-learn & Flask.
Built GenAI-powered solutions using LLMs and Hugging Face, reducing manual processing by 50%.
Developed cloud-hosted applications on AWS and Vector DB for internal clients.
Automated data pipelines and dashboards, cutting report generation time by 75%.

Apr 2022 – Sep 2022

Innodatastics

Data Science Intern

Hyderabad, India

Performed extensive data cleaning, visualization, and feature engineering on real-world datasets.
Built ML models for classification and regression (achieved 85%+ accuracy in test environments).
Supported model evaluation and improvement using cross-validation and hyperparameter tuning.

Capabilities

Skills

Stack used to deliver production RAG, LLM fine-tuning, and end-to-end ML systems.

Languages & Frameworks

Python92SQL82Scikit-learn90TensorFlow78Pandas92NumPy92Matplotlib85Seaborn82Flask84FastAPI86HTML / CSS78JavaScript72

Machine Learning & AI

Machine Learning (ML)90Deep Learning (DL)82Natural Language Processing (NLP)86Neural Networks80Computer Vision84Generative AI88Large Language Models (LLMs)86Hugging Face85Fine-tuning82RAG (Retrieval Augmented Generation)88Agentic AI80OpenCV82

Tools & Platforms

Jupyter / Google Colab92Git / GitHub85Docker72VS Code90Pinecone84Qdrant DB86AWS74Azure70OpenAI API88Sentence Transformers84ML Pipelines / CI-CD78Data Cleaning & Feature Engineering90Tableau78Postman75

Proof