Skip to content
View cherry51015's full-sized avatar

Block or report cherry51015

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
cherry51015/README.md

Typing SVG


Email  LinkedIn  McKinsey Forward


I build systems that close the gap between LLM capability and production reality. Second year CS undergrad at IIIT Kota — shipped 17+ GenAI applications across legal intelligence, multi-agent translation, RLHF pipelines, and financial multimodal RAG.

Currently thinking about: multi-agent memory architectures, MCP integrations, and what RAG reliability actually requires at scale.



↗ Projects

Levi — Legal Document Intelligence

153K+ clause FAISS index. Multilingual semantic Q&A, compliance verification, automated briefing. FastAPI + Gemini + GCP Cloud Run. 89% answer relevance, sub-200ms latency. Hallucination guardrails via strict document-grounding.

FastAPI FAISS Gemini GCP RAG

Agent Translator — Multi-Agent Translation

Hierarchical agents translating academic documents while preserving equations, tables, and layout — not just text. Self-correcting QA loop. 96% BLEU. 70% cost reduction vs. commercial alternatives.

CrewAI GPT-4 AWS ECS Multi-Agent

NarrativeForge — RLHF Content Pipeline

RAG + preference-based reward modeling for long-form generation. Built the feedback loop, not just the model. 34% hallucination reduction through multi-agent orchestration.

LangChain ChromaDB FastAPI RLHF

FinSight — Multimodal Financial RAG

KPI extraction and chart understanding from financial reports. 91% numerical accuracy. Serverless architecture on AWS Lambda.

Amazon Bedrock Claude Lambda RAG


⚙ Stack



LangChain LangGraph CrewAI AutoGen LlamaIndex FAISS Pinecone HuggingFace Neo4j MCP


◎ Background

ML Engineer — Fox Trading  Oct 2024 – Jan 2025
Feature pipelines over 500+ daily market indicators. LSTM + XGBoost ensemble at 85% directional accuracy on forex price prediction.

Open Source — LangChain
Contributed to RAG guardrails, retrieval refinement, backend optimization.

Technical Writer & Workshop Facilitator — QnS Club, IIIT Kota
12+ ML/engineering articles · 600+ readers · Delivered advanced prompt engineering sessions to 60+ students.


◈ Recognition

McKinsey Forward Scholar  ·  Stanford ML Specialization  ·  Google × Kaggle GenAI Intensive  ·  Faculty Pick, Vogue Vista AI  ·  LangChain Contributor



  



Pinned Loading

  1. Levi-legal_AI_assistant Levi-legal_AI_assistant Public

    This is production-grade legal AI assistant with 153K+ clause FAISS index enabling multilingual semantic Q&A, compliance verification, and automated briefing generation. Architected modular pipelin…

    Python

  2. LangchainProjects LangchainProjects Public

    A hands-on repository of LangChain-based projects and experiments, covering LLM integration, RAG workflows, and vector databases using langGraph, Streamlit, Ollama, AstraDB, HuggingFace, FAISS, Chr…

    Jupyter Notebook