About

I am Daniel C Cherian, a B.Tech graduate in Computer Science (AI & Machine Learning) from Sahyadri College of Engineering & Management, Mangaluru, set to graduate in 2026. During my studies, I completed internships at TrueByte Innovation LLP and Revstar Solutions, where I built backend systems, AI chatbots, and RAG pipelines using Python, FastAPI, and LangChain. My strongest skills include backend development, AI application building, and cloud deployment using Docker and AWS. I am eager to take on my first full-time role as a Software Engineer, contributing to engineering teams with my technical expertise and passion for innovation.

Experience

1. Software Engineer Intern

   TrueByte Innovation LLP

   Jan 2026 – May 2026

   Built a microservice backend using Python, FastAPI, and PostgreSQL, fully secured with JWT, BCrypt, and RBAC. Optimized database performance using joined-queries, and prevented race conditions by using DB-level ID allocation for bulk data imports. Automated deployments using Docker and a CI/CD pipeline, and implemented a two-phase commit pattern for reliable cloud file storage.

2. Product Intern

   Revstar Solutions

   Jul 2025 – Aug 2025

   Built an multi model AI chatbot and end-to-end RAG pipeline in Python to enable document-based Q&A. Implemented an end-to-end document intelligence pipeline using OCR and a persistent VectorDB to extract and retrieve data from scanned documents. Conducted systematic testing and debugging across pipeline stages to ensure production-ready reliability.

Education

1. Sahyadri College of Engineering & Management, Mangaluru

   Bachelor of Engineering in Computer Science (AI & Machine Learning)

   2022 – 2026 · Grade: 7.4/10

Projects

• Multi-Agent RAG System

  • Python
  • FastAPI
  • PostgreSQL
  • pgvector
  • Langgraph
  • Langchain

  Built an Agentic RAG system using 16-stage architecture pipeline using LangGraph for decision making. Where queries are secured, cached and passed through pgvector and BM25 and a 4-layer reranker. Before generating, CRAG grades, If it’s irrelevant a multi-agent loop takes over using ReAct, PlanExecute and RAPTOR for hierarchical retrieval. Before answer generated, Self-RAG repairs unsupported claims, followed by automated evaluations using RAGAS and LLM-as-Judge in the background.

• Real-Time Sleep Apnea Detection using TinyML and ECG

  • TensorFlow Lite
  • Python
  • CNN

  Designed an cost-effective, real-time Edge AI detection system by preprocessing the MIT ECG dataset to remove the noise and extracts critical heart rate features. Trained and optimized a CNN that successfully detected apnea events with a 94% accuracy and 0.98 ROC-AUC score. Deployed the model directly onto an ESP32 microcontroller using INT8 quantization to compress the data, enabling real-time, offline monitoring with millisecond latency.

• Full-Stack AI Assistant

  • React
  • FastAPI
  • Python
  • LangChain
  • LangGraph
  • FAISS
  • Ollama
  • RAG

  Built an AI assistant using LangGraph and LangChain with an agentic multi-step reasoning pipeline and multi-model LLM routing. Implemented a hybrid RAG pipeline with reranking to fetch and rank relevant document context. Added automated hallucination detection to verify answers before response.

Certifications

• Oracle Cloud Infrastructure 2025 Certified Generative AI Professional

  Oracle · 2025

• Career Essentials in Generative AI

  Microsoft & LinkedIn

• Career Essentials in Cybersecurity

  Microsoft & LinkedIn