// Compute Graph · Node 01 · pritamchaurasiya.com.np

Pritam Chaurasiya

AI Software Developer · LLM Agents

M.Sc. candidate in Data Science (AI/ML) at FAU Erlangen-Nürnberg and Working Student — AI at Siemens Healthineers, architecting agentic AI systems for healthcare log analysis, domain-specific RAG pipelines and language-model evaluation. I combine production-focused AI engineering with full-stack development using Python, FastAPI, JavaScript and MongoDB, backed by enterprise experience in Oracle APEX, PL/SQL and BPM automation.

Portrait of Pritam Chaurasiya

// System Summary

Runtime Profile

3+
Years building software — from enterprise ERP to AI systems
6+
Projects shipped — RAG, task automation, ERP workflows
4
Databases in production — Oracle, MySQL, MongoDB, PostgreSQL
2
Countries — India & Germany, remote and on-site delivery

// Modules

Capabilities

Six core modules, wired together into end-to-end systems — from raw documents and meeting audio to deployed, automated applications.

MOD-01

LLM Agents & RAG

Agentic pipelines with LangChain and LangGraph — retrieval-augmented generation, tool use, and context-aware reasoning over private documents.

LangChainLangGraphGemini APIPrompt Engineering
MOD-02

Full-Stack Development

RESTful backends and live web frontends — real-time task tracking, collaborative interfaces and modular APIs built for scale.

FastAPIJavaScriptMongoDBHTML/CSS
MOD-03

NLP & GenAI

Text and audio intelligence — extraction of actionable tasks from meeting transcripts, semantic search, summarization and question answering.

NLPEmbeddingsSemantic SearchGenAI
MOD-04

Enterprise Integration

ERP-connected systems on Oracle APEX and PL/SQL — REST/SOAP services, multi-stage BPM/BPEL approval flows, cross-database sync.

Oracle APEXPL/SQLREST/SOAPBPM/BPEL
MOD-05

Workflow Automation

Event-driven automation that removes manual work — webhooks, n8n pipelines and BPM tooling connecting Dropbox, Fireflies and ERP systems.

n8nWebhooksBPMAgile/Scrum
MOD-06

Data Engineering

Reliable data layers behind every application — schema design, ingestion and embedding pipelines, MySQL ↔ Oracle synchronization.

Oracle SQLMySQLPostgreSQLAWS

// Recent Work

Featured Projects

P-01 · LLM System

AskMyDocs — Document Intelligence

A Retrieval-Augmented Generation system for intelligent querying over user documents. Built the ingestion and embedding pipeline, integrated Gemini APIs for Q&A, summarization and semantic search, and tuned retrieval logic to cut hallucinations. Modular backend designed for future API and enterprise integration.

PythonLangChainRAGGemini
P-02 · Full-Stack AI

Taskflow — AI Task Automation

A task-management platform that extracts actionable tasks from meeting audio and text using NLP pipelines and LLMs. FastAPI backend with MongoDB for real-time storage, webhook integrations with Dropbox and Fireflies, and a collaborative web interface for live project tracking.

FastAPIMongoDBNLPWebhooks
P-03 · Enterprise

Oracle ERP Workflow Automation

Customer-onboarding and workflow automation systems on Oracle APEX — REST/SOAP integration with Oracle ERP, multi-stage BPM/BPEL approval flows, and MySQL ↔ Oracle database synchronization. Reduced manual intervention and improved process efficiency across business teams.

Oracle APEXPL/SQLBPM/BPELREST/SOAP

// Execution Log

Work Experience

08/2026 – PresentForchheim, DE

Working Student — AI

Siemens Healthineers

  • Architecting an AI agent to analyze complex diagnostic logs from medical systems during remote update cycles.
  • Building custom RAG pipelines and fine-tuning language models on domain-specific healthcare machine logs to surface actionable troubleshooting insights.
  • Designing test suites and benchmarking model performance to optimize retrieval accuracy, latency and diagnostic reliability.
08/2025 – 10/2025Nürnberg, DE

Working Student — AI Software Developer

SunSniffer GmbH & Co. KG

  • Developed AI applications using Python, LangChain and LangGraph for document automation.
  • Built AskMyDocs, an LLM-based RAG system for semantic search and document Q&A.
  • Developed a full-stack task-management system with FastAPI, JavaScript and MongoDB enabling real-time project tracking.
  • Improved internal productivity workflows through automation tools; agile, sprint-based delivery.
06/2023 – 03/2025Gurugram, IN

Senior Executive — Applications

SRF Limited

  • Developed enterprise cloud applications using Oracle APEX, PL/SQL and JavaScript.
  • Integrated REST/SOAP web services with Oracle ERP systems.
  • Automated business flows with Oracle BPM & BPEL tooling.
  • Optimized legacy codebases, improving runtime efficiency by ≈30%.
  • Provided cross-team technical support and led feature deployments.
01/2023 – 06/2023Gurugram, IN

Internship Trainee — Applications

SRF Limited

  • Developed internal modules with Oracle APEX and PL/SQL; built backend APIs and database procedures.
  • Automated approval workflows using BPM tools and integrated systems via REST and SOAP services.
  • Supported testing, debugging and feature deployment.

// Training Data

Education & Stack

04/2025 — Present

M.Sc. Data Science (AI/ML)

FAU Erlangen-Nürnberg, Germany
Focus: Machine Learning, LLMs, Generative AI, Data Engineering, NLP
Coursework: Deep Learning, Machine Learning, Data Science, AI Systems

07/2019 — 06/2023

B.Tech Computer Science & Engineering

Vellore Institute of Technology, India — CGPA 8.3
Focus: Software Engineering, Databases, AI, Cloud Systems

AI & GenAI

LLM Agents (LangChain, LangGraph) · RAG · NLP · Python AI Stack

Programming

Python · Java · JavaScript · PL/SQL

Web & Integration

FastAPI · Oracle APEX · REST & SOAP APIs · HTML/CSS

Databases

Oracle Database (SQL) · MySQL · MongoDB · PostgreSQL

Tools & Cloud

AWS · Git · n8n · Workflow Automation · JDeveloper (BPM/BPEL) · Agile (Scrum)

Languages

English (C1) · German (A1 — Active Learner)

// Handshake

Let's build something intelligent.

Open to working-student roles, AI/full-stack projects, research collaborations and professional connections. The fastest route is email — response latency is low.