AI QUARTERLY
PRASANNA NAGARALE · VOLUME I · 2026
Going to press…
0%
PRASANNA NAGARALE QUARTERLY · VOL. I · N° 001 · AI & MACHINE LEARNING · VIJAYAPURA, INDIA · 2026
COVER STORY

PRASANNA NAGARALE

AI  /  ML  ENGINEER

"I don't just use AI.
I build the systems that think."

Production AI · RAG Pipelines · LLM Fine-tuning · Multi-Agent Platforms

AI Quarterly — Prasanna Nagarale Cover
SCROLL
01

THE
ENGINEER

PROFILE & BACKGROUND

I am Prasanna Nagarale — an AI/ML Engineer from Vijayapura, Karnataka, India. I build AI systems that solve real-world problems at scale.

Completed a B.E. in Artificial Intelligence & Machine Learning at Basaveshwar Engineering College (BEC) in 2025. My work spans GenAI, NLP, Computer Vision, and full-stack AI deployment.

I've shipped platforms serving 500+ real users, published models on Hugging Face, and built systems hitting 99.3% accuracy in production. I thrive at the intersection of research and engineering — turning complex AI papers into tools that actually work for real people.

Open to AI/ML opportunities
Turning complex AI papers into tools that actually work for real people.
— Prasanna Nagarale, AI/ML Engineer

EDUCATION

B.E. in AI & Machine Learning
Basaveshwar Engineering College
2021 — 2025

CERTIFICATIONS

  • Career Essentials in GenAI — Microsoft & LinkedIn
  • Machine Learning with Python — IBM
  • NLP — Infosys Springboard
  • AI Fundamentals — Infosys
0AI Projects Built
0Live Users Served
0ML Accuracy
0LLM Speed Gain
0HuggingFace Models
0Internships
02

THE
WORK

PROJECTS & SYSTEMS BUILT
LIVE IN PRODUCTION
GENERATIVE AI · PRODUCTION SYSTEM wordriftai.com ↗

The AI Content
Platform
That Ships.

WordRift AI is Prasanna's flagship production system — a live AI content platform serving over 500 real users at 92% accuracy. Built end-to-end with Django, FastAPI and Transformers, it achieved a 75% reduction in LLM response time through intelligent caching and optimisation strategies.

DjangoFastAPITransformersPostgreSQLLLMs
500+Users
92%Accuracy
75%Faster
Visit Live Site →
RAG · AGENTIC AI

QueryFox AI

Advanced RAG system with intelligent agentic routing. Combines retrieval-augmented generation with agentic web search for document intelligence at scale.

RAGLangChainVector DB
GitHub →
🤗 HF PUBLISHED · PEFT

GPT-2 QLoRA Summarizer

Fine-tuned GPT-2 using QLoRA (PEFT) for abstractive text summarisation under limited compute. A study in parameter-efficient fine-tuning.

GPT-2QLoRAPEFT
HuggingFace →
🤗 HF PUBLISHED · NLP

Phi-2 Financial Sentiment

Fine-tuned a 2.7B parameter Phi-2 LLM using QLoRA for financial sentiment analysis. Publicly available on Hugging Face for the research community.

QLoRAPhi-2Fine-tuning
HuggingFace →
MULTI-AGENT · EDUCATION

Math Mentor AI

Multimodal multi-agent mathematics tutor for JEE-level problems using RAG and 5+ specialised agents orchestrated via Groq for adaptive reasoning.

RAGMulti-AgentGroq
GitHub →
DOCUMENT AI · MULTI-AGENT

AI Medical Claim Processor

FastAPI + LangGraph multi-agent system for processing PDF medical claims with intelligent document segregation and automated review pipelines.

LangGraphFastAPIPDF AI
GitHub →
COMPUTER VISION · IMAGE AI

HDR Image Enhancement

Computer vision pipeline for professional HDR image processing. Achieves 90% quality output in just 2 seconds — speed and quality engineered in tandem.

OpenCVComputer VisionPython
GitHub →
DOCUMENT AI · RAG

DocuMind RAG

Intelligent document Q&A using an advanced RAG pipeline with semantic search, reranking, and FAISS vector database for precise answers.

RAGFAISSLangChain
GitHub →
99.3% ACCURACY · CORE ML

Crop Recommendation

Production ML system for 22 crop types achieving 99.3% accuracy using Random Forest, deployed on Flask. Precision engineered without over-complexity.

Random ForestFlaskScikit-Learn
GitHub →
HEALTHCARE AI · CLASSIFICATION

Alzheimer's Detection

Medical AI for early Alzheimer's risk detection — analysing clinical, cognitive and lifestyle data. Applied AI where stakes and human impact are highest.

Medical AIClassificationPython
GitHub →
View all work: github.com/prasanna-nagarale ↗ · 14+ projects and counting
03

THE
STACK

CAPABILITIES REPORT

Six domains of expertise — built in production, not in tutorials.

I
LLM & GenAI
The Foundation
LLMsRAGFine-tuningQLoRAPEFTPrompt EngineeringHugging Face
II
Agent Frameworks
Orchestration Layer
LangChainLangGraphAutoGenMulti-AgentFAISS
III
Core ML & Data
The Engine
PythonPyTorchTensorFlowScikit-LearnNumPyPandas
IV
Backend & APIs
Production Layer
FastAPIDjangoFlaskREST APIsPostgreSQLMongoDB
V
CV & NLP
Perception Layer
OpenCVNLTKspaCyTransformersSentence BERT
VI
DevOps & Tools
The Scaffold
DockerGitGitHub ActionsJupyterStreamlitVS Code
04

THE
RECORD

PROFESSIONAL EXPERIENCE

Three chapters of professional growth — from web foundations to production AI at scale.

OCT 2025
→ APR 2026
MOST RECENTINTERNSHIP

AI Intern — GenAI + Prompt Engineering

AIVariant · Remote
  • Intensive internship on Generative AI and advanced Prompt Engineering
  • Worked extensively with ChatGPT APIs and production LLM systems
  • Applied QLoRA and PEFT for parameter-efficient model fine-tuning
GenAIPrompt EngineeringLLMsQLoRA
FEB 2025
→ JUL 2025
INTERNSHIP

Software Dev Intern — AI/ML Specialisation

ClickCrawl Media Pvt. Ltd. · Hubli, Karnataka
  • Engineered WordRift AI — production platform serving 500+ users with 92% accuracy
  • Reduced LLM response time by 75% through model optimisation and intelligent caching
  • Built HDR image enhancement system achieving 90% quality in 2 seconds
  • FastAPI backend with PostgreSQL and real-time analytics dashboard
DjangoFastAPITransformersOpenCVPostgreSQL
FEB 2023
→ MAR 2023
INTERNSHIP

Web Development Intern

FUEL · Hubli, Karnataka
  • Full-stack web development using HTML5, CSS3 and Python
  • Built responsive applications in a professional team environment
HTMLCSSPython
05

THE
SIGNAL

CONTACT & CONNECT

Transmit
your signal.

Open to full-time AI/ML roles, freelance projects, research collaborations, and interesting problems worth solving. Response time: under 24 hours.

Currently open to opportunities
Send a messageAll fields required