s
soumyajitsensha

Soumyajit

@soumyajitsensha
India
English, Hindi
About me
Hey there! I’m Soumyajit, an AI Engineer with hands-on experience creating intelligent systems that power real-world applications. From automating government patent workflows to building voice-enabled RAG chatbots, I bring end-to-end AI solutions to life. Agentic AI workflows using CrewAI, AutoGen, LangChain Multilingual summarization pipelines using mBART, BART 3D design generation with Stable Diffusion & Hunyuan3D Advanced semantic search and caching for fast AI responses Integrated AI tools into web backends using FastAPI + Docker ... Read more

Skills

s
soumyajitsensha
Soumyajit
Offline • 
Average response time: 1 hour

See my services

AI Integrations
I will build custom ai chatbots and agents using langchain and openai
From$100 / projectMore details
AI Integrations
I will build ai agents and automation systems
From$90 / projectMore details

Portfolio

Work experience

AI Engineer

Trademarkia • Full-time

Dec 2024 - Present • 1 yr 5 mos

AI Engineer building GenAI solutions Developing Agentic Solution for automating CRM communications Developing FastAPI application for AI Automation services

AI Research Intern

Samsung R&D • Full-time

Aug 2024 - Jun 2025 • 10 mos

At Samsung, I worked on developing an AI-powered agent for home automation that could understand natural language commands and intelligently control smart devices. The agent was designed to interpret user intent, map it to device-specific actions, and execute operations dynamically across different home appliances. I designed the system to allow device onboarding through APIs, where each device was described by its functionality, usage patterns, and live status. The agent analyzed these descriptions along with real-time device data to decide the correct action for a given user command. This enabled flexible control without hardcoding device logic. The project involved building agentic decision-making logic, integrating LLM-based intent understanding, and implementing action planning and execution workflows. I focused on making the system scalable so that new devices could be added with minimal configuration. Overall, the project demonstrated how AI agents can orchestrate real-world systems, combining natural language understanding, reasoning, and API-based execution to create an intelligent, extensible home automation platform.