Hi, I'm

Vikranth Bandaru

AI/ML Engineer & Data Scientist

Building real-world, scalable generative AI solutions that solve complex business problems. Specializing in LLMs, NLP, RAG systems, and full-stack data product development.

Vikranth Bandaru

About Me

I build AI systems that don’t just demo well, they ship, scale, and solve real business problems.

Recently graduated with an M.S. in Artificial Intelligence from the University at Buffalo, I’ve shipped production ML systems across data science, ML engineering, and AI product roles.

LLM & RAG Systems

Production-ready RAG pipelines with Llama models, vector databases, and scalable inference stacks focused on latency and reliability.

ML at Scale

Large-scale forecasting for 6K+ SKUs at 86%+ accuracy. End-to-end pipelines from data ingestion through monitoring and iteration.

AI Products

“Data Structurizer,” an OCR + LLM document intelligence engine that cut processing costs by 15%. Built to create measurable business value.

Tech Stack

Python, PyTorch, PySpark, SQL, LangChain, Azure/GCP, MLflow. IEEE-published researcher in computer vision.

Thinking deeply about applied AI? Let’s connect.

3+ Industries Served
10+ Deployed AI Models
50% Efficiency Boost
10M+ Data Points Analyzed

Technical Skills

Languages & Databases

Python
R
JavaScript
C/C++
SQL
MongoDB
PostgreSQL
Firebase
Qdrant Qdrant
Pinecone Pinecone

AI & Machine Learning

LLMs LLMs
NLP NLP
Deep Learning Deep Learning
PyTorch
TensorFlow
Keras
Scikit-learn
HuggingFace HuggingFace
LangChain LangChain
RAG RAG

Cloud & Tools

GCP
Azure
Docker
Git
Linux
PySpark
Vertex AI Vertex AI
Document AI Document AI
Power BI Power BI
n8n n8n

Frameworks & Libraries

React
Node.js
Flask
Streamlit
Pandas
NumPy
SpaCy
NLTK NLTK

Professional Experience

AI Engineer Intern

Vaspian, LLC

Sep 2024 – Dec 2025

  • Developed an AI-powered FAQ automation RAG system using LLMs (Llama 3.3 70B, Llama 4 Scout) and Qdrant to convert 12,000+ raw call transcript JSON files into structured problem–solution knowledge
  • Reduced customer problem-resolution time by 50% (from 6 minutes to 3 minutes) through instant, accurate chatbot answers
  • Integrated MCP servers and Dockerized services with Git-based version control, improving deployment reliability and iteration speed
  • Built an ElevenLabs conversational AI agent linked to Airtable via webhooks, automating webinar invites and customer response tracking

Data Scientist

Genpact

Jul 2023 – Aug 2024

  • Engineered ML pipeline for "Price-Pack Architecture" using PySpark, LLMs, and elasticity modeling to simulate over 1,000 pricing scenarios across 25+ markets and 6,000+ SKUs for Nestlé Purina, achieving 86% forecast accuracy
  • Developed Data Structurizer for invoice query processing: automated OCR extraction, performed ETL, stored in vector DB, and utilized GPT-4o API with LangChain for efficient invoice search, achieving 80% query-response accuracy
  • Built self-serve invoice Q&A chatbot for Genworth and Unilever using GCP Document AI, enabling rapid, accurate answers from invoice content and reducing manual lookup effort

Software Engineering Intern

Genpact

Mar 2023 – Jul 2023

  • Designed and deployed a web-based mentor-mentee platform for Hopeworks NGO using React, SQL, and Python, increasing matching engagement by 30%
  • Led development of a full-stack legal services platform using React, Node.js, MongoDB, and Firebase, integrating versioned docs, authentication, chat, and analytics

ML Engineer

HighRadius

Jan 2022 - Apr 2022

  • Built and deployed an AI-enabled fintech cloud application with React, MongoDB, and Python-based ML models (XGBoost, Random Forest), improving efficiency by 30%
  • Transformed raw, unstructured invoices into structured data, engineered features, and implemented ML models to identify the true invoice date among multiple fields

Education

Master's, Artificial Intelligence

University at Buffalo - SUNY

Aug 2024 - Dec 2025

Relevant Coursework: Machine Learning, NLP, Reinforcement Learning, Pattern Recognition, Info Retrieval, Data Intensive Computing, Analysis of Algorithms

Bachelor's, Computer Science (AI/ML)

SRM Institute of Science and Technology

Jun 2019 - May 2023

Relevant Coursework: Python, Artificial Intelligence, Applied Machine Learning, Statistical Machine Learning, NLP, Digital Image Processing, Computer Architecture, Advanced Calculus, Probability and Queueing Theory, Operating Systems, DBMS, Compilers

  • Awarded 100% tuition waiver for my Bachelor's degree on Merit Scholarship basis (SRMJEE All India Rank 270)
  • Graduated with GPA: 3.84/4.00 (9.14/10)
  • IEEE-published research on person re-identification

Honors & Awards

Best Dedication & Persistency Award

Best Dedication & Persistency

Genpact YoDA Recognition | Dec 2023

Honored with the Best Dedication & Persistency award as part of Genpact’s "Best Yoda" recognition, celebrating unwavering resilience, consistent effort, and a never-give-up spirit in the face of challenges. This award recognized my commitment to going above and beyond, fostering a culture of perseverance, and contributing meaningfully to the team’s success.

Bright Beginner Award

Bright Beginner

Genpact Futurero Internship Program | Jun 2023

Received the Bright Beginner award during Genpact’s prestigious Futurero Internship Program, recognizing my performance, adaptability, and learning mindset early in my professional journey. This recognition honors individuals who stand out by overcoming steep learning curves, embracing new challenges with curiosity and grit.

Testimonials

What people say about working with me

Featured Projects

Cognify - Fully Local Document Intelligence

Cognify - Fully Local Document Intelligence | PageIndex Retrieval | Zero Cloud Dependencies

Cognify is a locally-run Q&A system that answers questions from two knowledge sources: a built-in library of 50,000 Wikipedia article summaries across 10 topics, and any documents you upload yourself (PDFs and web pages). It runs entirely on your machine with a local LLM via Ollama. No API keys, no cloud subscriptions, no internet connection needed after the initial model download.
Python Retrieval-Augmented Generation (RAG) Large Language Models (LLM) PageIndex CI/CD Application Packaging
AI Code Reviewer - Production-Grade AI Security Analysis

AI Code Reviewer

A production-grade AI-powered code review system that automatically analyzes pull requests for security vulnerabilities, bugs, performance issues, and code quality problems.


Features:
  • Security Analysis: SQL injection, XSS, command injection, hardcoded secrets
  • Bug Detection: Null pointer dereferences, race conditions, logic errors
  • Performance: N+1 queries, memory leaks, inefficient algorithms
  • Dependency Scanning: CVE detection via OSV API
  • AI-Powered: Uses LLMs for deep code understanding
  • Multi-Language: TypeScript/JavaScript, Python, Go
Python OpenAI GitHub Actions DevSecOps Automated Testing Docker
SecureDoc - Professional Document Redaction Add-in

SecureDoc - Professional Document Redaction Add-in

A powerful Word Add-in that automatically redacts sensitive information, adds confidentiality headers, and enables tracking changes to maintain document security and compliance.


What it does:
  • Redact Sensitive Information: Retrieve document's complete content, locate sensitive information (emails, phone numbers, SSNs), and replace with redaction markers.
  • Add Confidential Header: Insert a "CONFIDENTIAL DOCUMENT" header at the top and ensure it is tracked by Tracking Changes.
  • Enable Tracking Changes: Use the Office Tracking Changes API to enable tracking changes.
TypeScript JavaScript CSS Vite Word JavaScript API HTML PowerShell Applied AI
LLMs in Social Sciences

LLMs in Social Sciences

Developed a multilingual policy analysis system using Python, Pandas, NLTK, DistilBERT, RoBERTa, and GPT-O2 to process English/Hindi sources. Achieved 87% accuracy in sentiment and stance detection for USA–India–Canada policy trends.
Python NLP DistilBERT RoBERTa Sentiment Analysis Selenium
Wikipedia Q&A Chatbot

Data-Driven Wikipedia Q&A Chatbot

Developed and deployed a chatbot using Python, Streamlit, RAG with OpenAI GPT, and Apache Solr to index 50K+ documents. Achieved 92% response accuracy with an interactive chat interface and analytics dashboard.
Python Streamlit RAG OpenAI GPT Apache Solr NLP
Social Distance Detector

Social Distance Detector

Developed an automated social distancing detection system using deep learning, tracking people in real-time video feeds to monitor and alert on social distancing violations.
Python Deep Learning Computer Vision OpenCV

Licenses & Certifications

Microsoft Certified: Azure Fundamentals

Microsoft

Issued Dec 2023 Credential ID: 534D76B6DAD80A37
Azure Cloud Fundamentals
Show Credential →

Generative AI

Genpact

Issued Dec 2023
Generative AI Artificial Intelligence (AI)
Show Credential →

Beginner Statistics for Data Analytics - Learn the Easy Way!

Udemy

Issued Apr 2023 Credential ID: 0004
Data Analytics Statistics Data Visualization Pandas ETL Microsoft Excel

Text Analysis and Natural Language Processing With Python

Udemy

Issued May 2023 Credential ID: 0004
Natural Language Processing (NLP) Python Scikit-Learn Pandas Text Analysis
Show Credential →

SQL and PostgreSQL: The Complete Developer's Guide

Udemy

Issued May 2023 Credential ID: 0004
SQL PostgreSQL Data Analytics Databases ETL
Show Credential →

Machine Learning | Natural Language Processing | Streamlit

Udemy

Issued May 2023 Credential ID: 0004
Machine Learning Natural Language Processing Streamlit Scikit-Learn Python
Show Credential →

Machine Learning, Data Science and Generative AI with Python

Udemy

Issued May 2023 Credential ID: 0004
Generative AI Data Science Machine Learning Python Scikit-Learn Pandas
Show Credential →

Programming for Everybody (Getting Started with Python)

Coursera

Issued Sep 2021 Credential ID: 8LJEZ22GTXVP
Python Algorithms
Show Credential →

Elements of AI

Reaktor By Google

Issued Sep 2021 Credential ID: jxb8d9x6cwh
Artificial Intelligence (AI)
Show Credential →

Publications

Person Re-identification from Video using Hybrid Approach

IEEE ICAECT 2023 | May 2023

This paper uses GANs and computer vision for person re-identification in video, improving matching accuracy in noisy settings. Published in IEEE Xplore and presented at the International Conference on Advances in Electrical, Computing, Communication and Sustainable Technologies.

Get In Touch

I'm currently looking for full-time opportunities in AI/ML and Data Science. Whether you have a question or just want to say hi, I'll try my best to get back to you!

Contact Information

Feel free to reach out through any of these channels

+1 469 439 0911
bandaruvikranth@gmail.com
New York, USA
OR
Schedule a Call