🧑‍🎓 I'm Ojasvi Goyal from Jaipur — a final-year B.Tech student at LNMIIT and alumni of DAV Public School, Kota, driven by curiosity.
🤖 Passionate about Agentic AI, Data Engineering, and NLP — building systems that reason, retrieve, and act.
🏢 Currently working as a Data Science Intern / Junior Associate AI Engineer at Celebal Technologies, contributing to production-grade multi-agent and data modelling systems.
🎮 Former GDG Domain Lead | Game design enthusiast | Project-driven learner.

🏢 Data Science Intern | Junior Associate AI Engineer — Celebal Technologies (Dec 2025 – Present)
On-site, Jaipur, Rajasthan

• Architected multi-agent orchestration frameworks using LangGraph for conversational reasoning pipelines — applied to retail inventory diagnostics (Eagle Eye) and autonomous transactional workflows (Agentic Commerce).
• Engineered a Graph-RAG Discovery Engine combining Neo4j and Azure OpenAI (GPT-4o) embeddings, with multi-hop Cypher queries and Databricks for structured evidence retrieval across complex schemas.
• Built an automated conditional purchase pipeline with Razorpay tokenization and background webhook catalogue scanners for real-time price-triggered transactions (Causal-X.AI).
• Implemented Zero-Trust security with OAuth 2.1, JWKS upstream token verification, and ECDSA P-256 (ES256) signatures for untampered agent-to-agent state serialization.
• Authored 48+ YAML test contracts for data quality diagnostics across Medallion architecture layers (Gold → Silver → Bronze → Raw) on the Eagle Eye platform.
• Contributed to the paradigm shift from conventional Medallion to ORDM (Open Retail Data Model) canonical architecture for retail use cases — via the Databricks Industry Solutions repository (Corporate ID: ojasvi-goyal-03).

🌐 Databricks Open Retail Data Model (ORDM) (Corporate: ojasvi-goyal-03)
Contributor to Databricks Industry Solutions. Worked on a paradigm shift from conventional Medallion to ORDM-based canonical and outcome packages — terminology and models oriented around real-world retail use cases.

🤖 Agentic Commerce & Causal-X.AI (Celebal Technologies)
Designed production-grade multi-agent architectures using LangGraph — covering conversational reasoning, autonomous transactional workflows, Razorpay tokenization for conditional purchases, and a Graph-RAG engine powered by Neo4j + Azure OpenAI.

👁️ Eagle Eye (Celebal Technologies)
Advanced data quality diagnostics platform. Authored 48+ YAML test contracts and implemented strict metadata ingestion validation across Medallion architecture layers for enterprise retail clients.

⚖️ LexNext.AI — Indian Legal Case Intelligence Platform
B.Tech Research Project | Under mentorship of Dr. Sarada Samantaray (LICAI - LNMIIT) | Ongoing
Processed 10,000+ Supreme Court cases. Built a semantic classification system using Sentence-BERT embeddings and a 50+ category legal taxonomy. Implemented Extractive (LexRank) + Abstractive (BART) summarization pipeline and Llama 3.1 8B for case-specific Q&A via RAG. Targeting a FastAPI + Next.js web interface.

🚦 Adaptive Traffic Control System
Fine-tuned YOLOv8 on urban Indian traffic datasets (IITM-HeTra) for real-time vehicle detection. Built a dynamic signal-timing engine in Python + Flask that adjusts green light durations based on live queue counts within a configurable ROI polygon.

📱 ContriKarle — Android Social App
Major Android project built with Kotlin, Jetpack Compose, MVVM architecture, and Firebase. Focused on clean UI/UX and real-time interactivity.

🎮 2D Game Development
Led the Game Dev vertical at GDG LNMIIT, mentoring juniors while building interactive 2D games with Godot and Unity, combining narrative, design, and physics.

"Hesitation is Defeat"