Hi, I'm

Shibu Mohapatra

AI Engineer

Building production-grade AI systems — LLMs, Agentic AI, Computer Vision, and Full-Stack solutions that scale.

Get in Touch View Projects

About Me

I'm an AI Engineer with 2.5+ years of experience building and deploying production-grade AI applications. My work spans Machine Learning, NLP, LLMs, Agentic AI, Computer Vision, and Full-Stack technologies.

I specialize in developing scalable AI systems, backend services, and automation workflows — with hands-on experience taking solutions from development all the way to production.

Currently focused on LLM-based applications, RAG pipelines, and Agentic AI systems that solve real-world problems at scale.

2.5+ Years Experience
350K+ MAUs Served
Top 10% Kaggle Ranking
9.8 M.Sc. CGPA

My Interests

AI & LLM Systems

Designing and shipping production-grade LLM applications — RAG pipelines, Agentic AI workflows, and real-time AI systems that solve real problems at scale.

Scalable Backend & Automation

Building robust APIs, async task pipelines, and microservices using Django, FastAPI, Celery, and Redis — and automating repetitive workflows end-to-end with RPA and AI tools.

Research & Open Source

Exploring emerging AI research — from computer vision and NLP to multi-agent systems — and contributing experiments, tools, and findings back to the community via GitHub and Kaggle.

Experience

AI & ML Engineer

Gloify
Dec 2024 – Present
  • Owned end-to-end development and deployment of an LLM-based personal care retail chatbot for Riyaah, integrating Django, prompt engineering, real-time streaming, Algolia product search, Pydantic Logfire, personalization workflows, image analysis, and a Vector DB-based RAG pipeline for FAQ retrieval.
  • Developed scalable AI backend infrastructure and APIs using MongoDB, AWS S3, Redis, Celery, RabbitMQ, JSON, and authentication services, leveraging Locust for load testing to improve reliability during high-traffic events.
  • Led QA teams on reliability testing and system validation for production AI deployments.
  • Optimized prompt workflows, built load testing scripts, and contributed to an ad-based subscription (monetization) feature for Agastyaa, an Agentic AI astrology assistant for AstroSure.ai, serving 350K+ MAUs within a 1M+ user base.
LLMsRAGDjangoMongoDBRedisAWS S3CeleryAgentic AI

Junior AI Developer

Skills Agency
Jun 2024 – Aug 2024
  • Developed an LLM-powered chatbot using LangChain and SQLAlchemy, enabling non-technical teams to access internal databases without writing SQL.
  • Deployed a WebRTC-based Gen AI chatbot using NLP, PyTorch, and a HeyGen avatar for conversational interaction.
LangChainNLPPyTorchWebRTCSQLAlchemy

RPA Engineer

Guidewire Software
May 2023 – Oct 2023
  • Collaborated with cross-functional teams across time zones to develop BI Finance Bots using Python, Machine Learning, and Automation Anywhere, automating GW dataset generation and saving 4+ hours of manual effort per report.
  • Implemented SoapUI for automated testing and RabbitMQ as a message broker to support reliable bot execution.
  • Integrated Oracle Fusion, Oanda foreign exchange, Vertex Global Tax, and Coupa P2P to streamline bot execution and improve revenue reporting accuracy.
PythonRPAAutomation AnywhereRabbitMQMachine Learning

Technical Skills

Languages

PythonRJavaScript

AI & ML

Data ScienceMachine LearningDeep LearningNLPLLMsRAGPrompt EngineeringGenerative AIComputer VisionOpenAI/GPTHugging FaceOllamaAgentic AI

Libraries

PandasNumPyscikit-learnTensorFlowKerasPyTorchMatplotlibOpenCVNLTK

Backend

DjangoFastAPIFlaskNode.jsMicroservicesCeleryRedisPydanticBeanieMotor

Databases

MySQLMongoDBChromaDB

DevOps & Cloud

DockerKubernetesAWSAzureMLOpsCI/CDJenkinsJiraGit

Automation

UiPathAutomation Anywheren8nAntigravityClaude

Tools

LinuxStreamlitAutoMLLangChainAlgoliaLogfireLocustPostmanRabbitMQSoapUISQLAlchemyTeamsSlack

Projects

Transcription

Multilingual real-time transcription app powered by Gemini for live speech-to-text across languages.

GeminiReal-timeMultilingual

Support Agent

AI-powered customer support agent for e-commerce, focused on intelligent user query processing.

LLMsAgentic AIE-commerce

War Room AI

Multi-agent system simulating a cross-functional war room during a product launch scenario.

Multi-AgentLLMsSimulation

ReportsAI

Multi-agent system that ingests a bug report and related logs, reproduces the issue, and outputs a root-cause hypothesis plus a patch plan.

Multi-AgentLLMsAgentic AI

Voice Scheduling Agent

Voice interactions to collect details, validate, and schedule calendar events using AI.

Voice AINLPAgentic AI

rPPG

Prototype that takes a 60-second face video as input and processes it in 5-second chunks to estimate heart rate using remote photoplethysmography.

Computer VisionOpenCVPython

AutoML.ai

An algorithm to automate an end-to-end ML pipeline using Streamlit and AutoML techniques.

AutoMLStreamlitML Pipeline

Education

Master of Science in Artificial Intelligence

University of Mumbai, India
Jul 2021 – Jun 2023 CGPA: 9.8 / 10

Bachelor of Science in Computer Science

Nagindas Khandwala College, India
Jul 2018 – Jun 2021 CGPA: 8.6 / 10

Achievements & Certifications

Competitions & Recognition

  • Kaggle — Top 10% ranking. Datasets Expert. Multiple medals in ML competitions.
  • DataHack — Click-Through Rate Prediction (74% accuracy). Forecast Hourly Energy Demand (RMSE). Vahan Bima Insurance (89% accuracy).
  • Blogs — ML model for phishing classification. NLP sentiment analysis model (94% accuracy). Benchmarking model for CNN, DNN, and LSTM on MNIST.

Certifications

Get in Touch

"The best AI systems don't just process data — they understand context, adapt to uncertainty, and deliver real-world impact."

mohapatrashibu@gmail.com +91 82082 13211 linkedin.com/in/shibu-mohapatra github.com/MohapatraShibu