I enjoy building things that bring together Artificial Intelligence, software engineering, automation, and real-world problem solving.
I'm a final-year B.Tech Artificial Intelligence & Machine Learning student at Sagi Ramakrishnam Raju Engineering College (SRKR), graduating in 2027.
My interests span from building autonomous AI agents and multimodal systems to engineering scalable APIs, developing full-stack applications, experimenting with machine learning, and exploring the mathematical foundations behind intelligent systems.
I don't just like experimenting with models — I like turning ideas into working, integrated, end-to-end systems.
- 🤖 Building agentic workflows, LLM-powered applications, and multimodal AI systems.
- 🧠 Exploring RAG, tool calling, memory architectures, AI reliability, and intelligent automation.
- ⚙️ Developing APIs, backend services, database systems, and full-stack applications.
- 👁️ Experimenting with computer vision, OCR, and visual automation.
- 📊 Interested in ML systems, data processing, and mathematical foundations of AI.
Agentic AI · Multi-Agent Systems · LLM Orchestration · RAG · Prompt Engineering · Tool Calling · Function Calling · Persistent Memory · LLM Evaluation · AI Automation · Gemini API · Groq · PhiData
OpenCV · EasyOCR · OCR Pipelines · Whisper · Image Understanding · Speech Processing · Multimodal Workflows · Visual Evidence Analysis
Python · Java · JavaScript · C/C++ · React.js · Node.js · Express.js · MongoDB · HTML · CSS · Tailwind CSS · FastAPI · Flask · REST APIs
Scikit-learn · Pandas · NumPy · XGBoost · Random Forest · Feature Engineering · Data Preprocessing · Model Evaluation · SQLite · SQL
Git · GitHub · Docker · Postman · VS Code · Jupyter Notebook · Google Cloud · Vertex AI · Cloud Storage · App Engine · NVIDIA GPU Computing
A collection of systems I've designed, developed, experimented with, and continuously improved.
An AI-powered analytics system that translates natural-language questions into SQL queries and executes them against structured databases.
The goal is not simply to generate SQL, but to build a more reliable query execution pipeline with validation, error recovery, and intelligent model orchestration.
What makes it interesting:
- Natural-language-to-SQL generation using LLMs.
- SQL AST validation to inspect and validate generated queries before execution.
- Query execution through a FastAPI backend with SQLite integration.
- Anomaly detection and query-result handling.
- LRU caching to reduce repeated processing.
- Multi-provider LLM fallback for improved resilience.
- Bounded self-healing workflows that attempt to recover from failed queries.
- API-first architecture with a Streamlit interface.
Tech: Python · FastAPI · SQLite · SQLGlot · LLMs · AI Agents · Streamlit · REST APIs
A multimodal AI workflow designed to process incoming information across text, images, and voice, understand the context, and route messages intelligently.
What makes it interesting:
- Text, image, and voice input processing.
- OCR-based extraction from images using EasyOCR.
- Speech-to-text transcription using Whisper.
- Context-aware message understanding.
- Persistent memory using a Hindsight-based memory pipeline.
- Deterministic routing logic with LLM-based fallback.
- Multimodal processing combined with automated decision-making.
Tech: Python · Agentic AI · EasyOCR · Whisper · LLMs · Multimodal AI · Persistent Memory · Workflow Automation
An AI-driven system focused on analyzing visual evidence and extracting meaningful, structured insights from multimodal inputs.
What makes it interesting:
- Multimodal visual evidence analysis.
- Structured evaluation of evidence.
- Output validation and consistency checks.
- Prompt-injection awareness and defensive processing.
- Caching and reusable analysis workflows.
Tech: Python · Multimodal AI · LLMs · Computer Vision · Prompt Engineering · Structured Outputs
A visual automation project that explores how computer vision and OCR can bridge the gap between cloud-based development environments and local project structures.
It identifies visible project structures and reconstructs them locally through automated visual interpretation.
What makes it interesting:
- Visual understanding of cloud IDE interfaces.
- OCR-based extraction of project and file information.
- GUI automation for interacting with development environments.
- Reconstruction of local directory and file structures.
- Experimentation with GPU-assisted processing.
Tech: Python · OpenCV · EasyOCR · PyAutoGUI · Computer Vision · GUI Automation · NVIDIA GPU
An LLM-powered financial research assistant that combines external tools and market data to produce contextual financial insights.
What makes it interesting:
- Integration of live financial data through YFinance.
- Web search for additional company and market context.
- LLM-powered reasoning and response generation.
- Tool integration for financial research workflows.
- Streamlit interface for interacting with the agent.
Tech: Python · PhiData · Groq · Llama · YFinance · DuckDuckGo · Streamlit · AI Agents
A full-stack fitness tracking application built around the MERN ecosystem, demonstrating my interest in application development beyond AI.
What makes it interesting:
- Full-stack application architecture.
- React-based frontend development.
- Node.js and Express backend development.
- MongoDB-based data management.
- CRUD operations and REST API integration.
- Connecting frontend interactions with backend services.
Tech: MongoDB · Express.js · React.js · Node.js · JavaScript · REST APIs
An AI-assisted application exploring how intelligent features can be integrated into a practical user-facing software experience.
Tech: AI Integration · Python · Full-Stack Development
A machine learning project focused on predicting credit card approval outcomes using structured data and supervised learning.
What makes it interesting:
- Data preprocessing and feature preparation.
- Random Forest classification.
- ML model integration with a Flask application.
- Model deployment and application-level prediction workflows.
Tech: Python · Scikit-learn · Random Forest · Flask · Pandas · Machine Learning
Worked on a data-oriented project involving sustainability-related datasets, preprocessing, structured storage, and visualization.
Areas explored:
- Data cleaning and preprocessing.
- SQLite-based data management.
- Data analysis and visualization.
- Applying computational techniques to real-world datasets.
Explored mathematical and computational aspects of higher-order matrix adjoints through implementation and experimentation in C.
Focus: Mathematical Computing · Matrix Operations · Linear Algebra · C Programming
- 🥇 Google The Big Code 2026: Qualified and progressed to the qualifier stage, among the top 15,000 participants nationwide.
- ⚡ HackerRank Orchestrate — June 2026: Ranked among the top 800 participants out of approximately 2,000.
- ⚡ HackerRank Orchestrate — August 2026: Ranked among the top 700 participants out of approximately 2,500.
- 🧮 National Ramanujan Mathematics Competition: Achieved Level 2.
- 💻 Smart Interviews SRKR-2027: Secured a position among the top 12 in Assessment 1 with 160 marks.
- 🤖 AI Internship: Shortlisted for an AI internship opportunity.
I'm particularly interested in the intersection of theory and practical AI engineering:
- Neural Networks and Computer Vision.
- Fullstack and MERN applications to production grade deployment
- Mathematical Foundations of Machine Learning.
- Machine Learning Systems and Data Processing.
- Agentic AI and Multi-Agent Architectures.
- Trustworthy, Reliable, and Efficient AI.
- Multimodal Learning and LLM-based Systems.
I'm always interested in connecting with people who are building, experimenting, researching, and solving interesting problems.
Whether it's an AI startup, an engineering challenge, an open-source project, a research idea, or a full-stack product — I'm excited to contribute, learn, and build.
- 💻 GitHub: Neer-tech878
- 💼 LinkedIn: Neeraj Kadali
- 🚀 Open to: AI/ML Engineering · Agentic AI · Full-Stack Development · Research · Startup Collaborations
"From intelligent agents to full-stack applications — I love turning complex ideas into working systems."
⭐ If something here interests you, feel free to explore my repositories!