Generative AI Engineer

Vikrant Vikaasa

I design RAG pipelines, multimodal systems, and agent-based workflows using Python, FastAPI, CrewAI, Google ADK, and modern LLM stacks. My recent work focuses on document intelligence, visual search, and customer support automation.

Current Role Software Engineer, Spritle
Focus RAG, Agents, Multimodal AI, Enterprise AI, Backend, Fine-tuning
Location Coimbatore, India
Hire Me GitHub
RAG
Agents + Vision
Portrait of Vikrant Vikaasa

About

Who I am and what I do.

I specialize in turning LLM capabilities into usable products: document search, extraction pipelines, multimodal understanding, and real-time tool-connected agents. My background in Artificial Intelligence and Data Science supports both the modeling side and the production engineering side.

Core Stack

Python FastAPI Streamlit Flask CrewAI Google ADK LangChain LangFlow Azure AI Services ChromaDB OpenAI GPT Qwen VLM Llama 3.2 SQL GitHub MCP Hugging Face Transformers Prompt Engineering

Education

B.Tech, Artificial Intelligence and Data Science KGISL Institute of Technology, Coimbatore 2021 - 2025 CGPA: 7.9
Higher Secondary Suguna RIPV Matric School, Coimbatore 2021 80%

Experience

Where I've worked and what I've built.

Jun 2025 - Present Chennai

Software Engineer (Gen AI Engineer)

Spritle Software

  • Designed and implemented Retrieval Augmented Generation (RAG) pipelines using LLMs to enable intelligent document search and conversational AI applications.
  • Built scalable backend services using FastAPI to support AI-powered applications for document processing, customer support automation, and data extraction
  • Developed multimodal AI systems integrating vision-language models to support visual search and intelligent document understanding.
  • Implemented vector search pipelines using ChromaDB for semantic document retrieval and similarity search across enterprise datasets.
  • Built agent-based AI workflows using CrewAI and Google ADK to automate complex reasoning tasks and multi-step query handling.
  • Integrated external data sources and tools using Model Context Protocol (MCP) to enable real-time access to databases and APIs within AI applications.
  • Delivered AI-powered solutions for enterprise use cases in HR and insurance domains, improving document analysis and customer interaction workflows.
Jul 2023 Coimbatore

Full Stack Developer Intern

Aroganam Technologies

  • Developed responsive web applications using HTML, CSS, JavaScript, and React.
  • Built and integrated frontend components with backend APIs for dynamic data rendering and interaction.

Projects

AI products and systems I've built.

My projects span enterprise AI solutions, multimodal search systems, agent platforms, and personal AI applications. I focus on leveraging LLMs, RAG techniques, and modern AI tools to build impactful products.

Enterprise AI Feb 2026

AI Powered Data Extraction

AI-powered extraction pipeline for Certificates of Analysis using Azure AI Content Understanding, Azure OpenAI GPT models, FastAPI, and Azure Blob Storage.

  • Automated structured extraction from PDFs and Excel documents.
  • Handled high-volume uploads and batch processing workflows.
  • Improved output consistency through prompt engineering.
FastAPI Azure AI Azure Blob OpenAI GPT
Visual Search Nov 2025

Visual Search

Product discovery platform using Azure OpenAI GPT and Qwen VLM to generate rich product descriptions and power similarity search over large catalogs.

  • Designed ChromaDB architecture for fast vector retrieval.
  • Built FastAPI services for real-time image processing and result delivery.
Qwen VLM Azure OpenAI ChromaDB FastAPI
Agent Platform Sep 2025

AI Agent Platform

E-commerce AI agent system built for intelligent query resolution and recommendation workflows with tool-connected, multi-model orchestration.

  • Implemented agent systems using LangChain, OpenAI, Mistral, and Agno agents.
  • Connected MCP tools for real-time database queries and external API access.
  • Applied prompt strategies for accurate, context-aware business responses.
LangChain Mistral API OpenAI API MCP
Personal Project Nov 2024

Git Chat

GitHub repository chatbot hosted on Hugging Face Spaces, built to answer questions over codebases using retrieval-augmented generation with an open-source Llama model.

  • Used GitHub access tokens to fetch repository content for question answering.
  • Built a Streamlit interface for interactive chat over repository context.
  • Powered responses with the Meta Llama 3.2 1B model in a lightweight hosted setup.
Streamlit RAG GitHub API Llama 3.2
Source Code
Personal Project Mar 2024

Interview Prep App

AI-powered platform that helps job seekers improve their resumes and interview skills. It provides AI-based resume analysis with personalized feedback and conducts mock interviews with tailored questions, sentiment analysis, and confidence scoring to enhance interview performance.

  • Analyzed resumes using AI to provide ATS scores and improvement suggestions.
  • Generated personalized mock interview questions based on resumes and job descriptions.
  • Integrated speech-to-text and AI feedback analysis for interview performance evaluation.
Flask ReactJs Gemini DeepSeek Groq Whisper Speech Recognition
Source Code
Personal Project Sep 2023

VoiceMG

Speech-to-text and grammar assistance application that records spoken input, transcribes it, and checks the resulting text for language quality.

  • Built with Streamlit for a lightweight interactive interface.
  • Used OpenAI Whisper for transcription and Llama 2 for grammar correction.
Streamlit Whisper Llama 2
Source Code

Contact

Open to AI product, backend, and applied ML roles.

If you are building with RAG, agents, multimodal search, or enterprise automation, I would be glad to connect.

vikrant0527@gmail.com +91 9787121207 linkedin.com/in/vikrantvikaasa