Adarsh Devarakonda
Summary
Confident and passionate AI Engineer with hands-on experience in agentic architectures, Retrieval-Augmented Generation (RAG), and custom LLM integrations. Skilled in building practical AI systems and automation workflows using Vibe coding, driven by curiosity and a strong desire to learn and solve challenging problems.
Overview
2
2
years of professional experience
1
1
Certification
Work History
AI Engineer
Sparity
03.2025 - Current
- SAP Support/Automation Platform: Engineered agentic ticketing and support workflows using Node.js/React.js, FAISS, LangChain, and custom LLMs (OpenAI, Llama 3.1, local models). Deployed agents for document parsing, SOP/KT doc generation, and full workflow automation in SAP BTP/ServiceNow.
- AI-Powered Dashboard Migration Tool: Developed a Tablue to Power BI migration solution with FastAPI/PostgreSQL. Built an auto-validation workflow leveraging GCP Vision API, OpenCV, Vision Transformers, YOLOv8, and OpenAI assistants for migration testing.
- News Digest Agents: Built multi-agent systems using FastAPI, Azure OpenAI, LangChain, FAISS; implemented article clustering, semantic search, comparative summaries, NER, and human-in-the-loop feedback.
Student Intern
SVNIT Surat
01.2024 - 05.2024
- Developed Crop Disease Detection solutions using CNNs for agricultural image analysis and precision yield prediction.
Education
B.Tech - Computer Science
SVNIT Surat
01-2024
Intermediate -
Kennedy High The Global School
Class X - undefined
NASR Boys School
Skills
- Languages, DB & Stacks: Python, Nodejs, React, FastAPI, SQL, PostgreSQL
- AI Tools: LangChain, CrewAI, FAISS, ChromaDB, Neo4J, HugginceFace, Unsloth, YOLOv8, Vision Transformers
- LLMs: OpenAI, Gemini, Groq, Claude, Azure OpenAI, Llama 31, custom local LLMs, prompt engineering and fine-tuning
- Agentic AI Systems: Multi-agent platforms, agent-linked knowledge models, supervisor agent orchestration using OpenAI assistants, Google ADK, A2A and MCP servers, RAG and Graph RAG Architecture with NER, semantic search and entity relations
- Vibe Coding & Cursor Mastery: Primary coding style; rapid prototyping and language independent coding across diverse AI projects
- Evaluation & Demos: Architected agent evaluation frameworks and model evaluation metrics; delivered internal office talks/demos on vibe coding and agent evaluation systems
- ML/DL: SciKit-learn, TessonFlow, Keras, Pandas, Numpy, Pytorch, OpenCV
Certification
- Agentic AI Frameworks (NASSCOM AI): Advanced agent-based modeling, MCP & MCP server integration, agent orchestration, Vibe Coding
- Generative AI with Large Language Models (DeepLearning.AI, AWS): Transformer architectures, LLM deployment strategies, prompt engineering
- Gen AI Intensive Course (Google/Kaggle): LangGraph, Gemini API, knowledge graph integrations, agent supervision
- Scaler Masterclass & AI Society Course: Advanced Python, ML/DL, CrewAI agent workflows
- Machine Learning with Python (IBM): Core ML algorithms, applied Python for ML workloads
- Vibe Coding Certification: Mastery in rapid agentic coding
Timeline
AI Engineer
Sparity
03.2025 - Current
Student Intern
SVNIT Surat
01.2024 - 05.2024
Class X - undefined
NASR Boys School
B.Tech - Computer Science
SVNIT Surat
Intermediate -
Kennedy High The Global School
Additional Skills
- Experience with GCP Vision/Vertex AI, SQL, Git, HTML/CSS.
- Strong foundation in ML, DL, Python, agentic system design.
- Effective communicator and presenter (office demos/talks on agentic coding and AI evaluation).