Located in Pune, India
Available for opportunities

Software & Analytics Engineer

Ashish Mali

Portrait of Ashish Mali

Data Engineer

Building data pipelines
with Databricks & Spark,
and AI applications with
RAG and agentic workflows.

Ashish Mali is a Software & Analytics Engineer based in Pune, India, currently a Project Student at Quorum Software, a product-based company, and a Project Trainee at Uplift Mutuals. He builds data pipelines with Databricks and Apache Spark, and AI applications with RAG and agentic workflows, while pursuing an MCA in Data Science at Savitribai Phule Pune University.

Screenshot of AI-Powered Resume Screening and Candidate Ranking System

AI-Powered Resume Screening and Candidate Ranking System

Generative AI · Research · Sep 2026 – Nov 2026

Semester III Research Project (RP641RP), Savitribai Phule Pune University

An explainable recruitment tool that reads a job description and resume PDFs, extracts each candidate's skills, experience and education with Google Gemini, and ranks candidates with a transparent scoring formula that shows exactly why each candidate got their score.

Key features
  • Job description pasted or uploaded (PDF, DOCX, TXT), plus bulk resume PDF upload with automatic text extraction
  • Semantic skill matching that understands synonyms and ignores keyword stuffing
  • Transparent weighted scoring (Skills 40%, Experience 30%, Relevance 20%, Education 10%) with adjustable weights
  • Matched skills, missing skills and a plain-English explanation for every candidate
  • Offline mode (keyword rules + TF-IDF) when no API key or internet is available
  • Screening history saved to a database with CSV export
  • Evaluated on 10 test resumes: 0.976 rank agreement vs 0.782 for a keyword-based (TF-IDF) baseline
  • Python
  • Streamlit
  • Google Gemini API
  • Pydantic
  • PyMuPDF
  • scikit-learn
  • SQLite
  • pytest
Screenshot of RAG-Based Telecom Customer Support Chatbot

RAG-Based Telecom Customer Support Chatbot

Generative AI · RAG

A RAG-based telecom customer support chatbot that delivers accurate, context-aware responses for billing, connectivity, SIM services, roaming, and other telecom-related issues.

Key features
  • Multi-source RAG pipeline for ingesting, embedding, and retrieving context
  • Support for 3+ diverse data sources to improve relevance and retrieval efficiency
  • Real-time conversational support with Groq-powered gpt-oss-120b and local LLMs
  • Streamlit interface for interactive customer support experiences
  • LangChain
  • ChromaDB
  • Hugging Face
  • Groq API
  • Docker Model Runner
  • Streamlit
Screenshot of Agentic AI Shopping Assistant

Agentic AI Shopping Assistant

Agentic AI

An agentic AI shopping assistant that enables natural-language product search, price and organic filters, rating lookup, and order placement through interactive conversation.

Key features
  • Natural-language shopping assistant for product discovery and filtering
  • Modular tool-based workflow for search, reviews, image analysis, and checkout
  • Persistent product and order data management with SQLite
  • Safe checkout testing with Groq vision and text models plus local LLMs
  • Streamlit
  • SQLite
  • Docker Model Runner
  • Groq API
  • Vision Models
  • Local LLMs
Screenshot of Loan Default Analysis & Prediction System

Loan Default Analysis & Prediction System

Machine Learning · Aug 2025 – Dec 2025

Semester End Mini Project (MP541MP), Savitribai Phule Pune University

An end-to-end ML web application for loan default analysis that performs data preprocessing, model training, prediction, and visualization through an interactive interface.

Key features
  • Loan default prediction with a random forest classifier
  • Interactive dashboards for data exploration
  • Secure user handling and data storage in MongoDB
  • End-to-end pipeline from data input to prediction output
  • Flask
  • MongoDB
  • scikit-learn
  • Plotly
  • Dash
Screenshot of Battle Prediction ML-PokeDex WebApp

Battle Prediction ML-PokeDex WebApp

Machine Learning · Web

Built with Atharv at Sri Balaji University, Pune

A Pokédex application with battle prediction using a random forest model, user authentication and detailed Pokémon information.

Key features
  • Secure login and registration
  • Pokédex with data for hundreds of Pokémon and advanced filtering
  • Random-forest battle outcome prediction
  • Python
  • Flask
  • Bootstrap
  • JavaScript
  • Supabase
  • REST API
Screenshot of DHS-NFHS Data Analysis

DHS-NFHS Data Analysis

Data Analysis

Analysis of health survey data from the DHS and NFHS datasets: data wrangling, exploratory data analysis, and visualization of health metrics to derive insights from large-scale survey data.

Key features
  • Large-scale health survey data processing
  • Exploratory data analysis and statistical insights
  • Interactive dashboards for health metrics
  • Python
  • Pandas
  • NumPy
  • Matplotlib
  • Seaborn
  • Plotly
Screenshot of Retail Data Engineering Pipeline for Point-of-Sale Analytics and Incentive Calculation

Retail Data Engineering Pipeline for Point-of-Sale Analytics and Incentive Calculation

Data Engineering · Mar 2026 – May 2026

Semester End Mini Project (MP581MP), Savitribai Phule Pune University

An automated ETL (Extract, Transform, Load) system that processes high-volume retail sales transactions from multiple store locations, performs real-time data validation and enrichment, and generates analytics for customer behaviour analysis and sales team performance evaluation.

Key features
  • Distributed data processing with Apache Spark
  • AWS S3 for cloud storage
  • MySQL for data warehousing
  • Real-time analytics and accurate incentive calculation
  • Apache Spark
  • AWS S3
  • MySQL
  • Python

Experience

  • Project student at Quorum Software, a product-based company. On-site in Pune, Maharashtra.
  • Project trainee at Uplift Mutuals. Remote, Pune, Maharashtra.
  • Four-month analytics engineering internship working with Amazon Web Services (AWS). Remote, Pune/Pimpri-Chinchwad.
  • Helped design, build and monitor 25% of the product's data pipelines.
  • Project data lead for an NFHS-V Indian diabetes analysis.
  • Six months on-site in Pune, Maharashtra.

Certifications & Patent

  • Oracle

    Oracle Cloud Infrastructure Certified AI Foundations Associate

    Verify credential ↗
  • Amazon Web Services

    AWS re/Start Graduate

    Verify credential ↗
  • Databricks

    Databricks Fundamentals

    Verify credential ↗
  • Databricks

    Databricks AI Agent Fundamentals

    Verify credential ↗
  • Kyoto Seika University · Sri Balaji University

    Special Japanese Language Course

    Three-week Japanese language and culture course (hiragana, katakana, numbers and traditions) taught at Sri Balaji University, Pune by exchange students from Kyoto Seika University. One of 60 students selected through a merit-based test; awarded a certificate of appreciation (感謝状) on completion.

    View certificate ↗

Blogs

All articles on Medium ↗

Skills & Tools

Programming

PythonSQLCC++

Data Engineering

DatabricksApache SparkPySparkETL Pipelines

Databases

PostgreSQLMySQLMongoDBChromaDB

AI & LLMs

RAGAI AgentsLangChainLangGraphPrompt EngineeringHugging Face

Frameworks

FastAPIFlaskStreamlitDashPydantic

BI & Visualization

Power BITableau

Tools

DockerGitGitHubLinuxKaggle

Frequently asked questions

Who is Ashish Mali?

Ashish Mali is a Software & Analytics Engineer based in Pune, India. He is currently a Project Student at Quorum Software, a product-based company, and a Project Trainee at Uplift Mutuals, and is pursuing an MCA in Data Science at Savitribai Phule Pune University (GPA 8/10).

What technologies does Ashish Mali specialize in?

Python and SQL; data engineering with Databricks, Apache Spark and PySpark; and generative AI with RAG, LangChain, LangGraph, ChromaDB and Hugging Face. He builds applications with FastAPI, Flask and Streamlit, and holds the Oracle Cloud Infrastructure AI Foundations Associate, AWS Certified Cloud Practitioner, Databricks Fundamentals and Databricks AI Agent Fundamentals certifications.

What is the RAG-Based Telecom Customer Support Chatbot?

A retrieval-augmented generation (RAG) chatbot that answers telecom customer questions about billing, connectivity, SIM services and roaming. It retrieves context from 3+ data sources with LangChain and ChromaDB, runs gpt-oss-120b through the Groq API or local LLMs through Docker Model Runner, and has a Streamlit interface.

What work experience does Ashish Mali have?

Since September 2026 he has been a Project Student at Quorum Software and a Project Trainee at Uplift Mutuals. Before that he was an Analytics Engineer Intern at Altus Cloud Software (June–September 2026), working with AWS, and a Databricks Data Engineer Intern at MyndNova Solutions (January–June 2026), where he helped design, build and monitor 25% of the product's data pipelines and was project data lead for an NFHS-V Indian diabetes analysis. He is also the inventor (software lead) on published patent application No. 202421016351 A.

What is the AI-Powered Resume Screening and Candidate Ranking System?

Ashish Mali's Semester III research project at Savitribai Phule Pune University: an explainable recruitment tool that extracts candidates' skills, experience and education from resume PDFs with Google Gemini and ranks them with a transparent weighted score (skills 40%, experience 30%, relevance 20%, education 10%). On 10 test resumes it reached 0.976 rank agreement, compared with 0.782 for a keyword-based TF-IDF baseline. It is built with Python, Streamlit, Pydantic, PyMuPDF, scikit-learn and SQLite.

How can I contact or hire Ashish Mali?

Use the contact form at ashishmali.me/contact or email lmno3418@gmail.com; he usually replies within 24–48 hours. He is also on LinkedIn and GitHub as lmno3418.