Skip to content
View smadinen7's full-sized avatar
4 followers · 4 following

Achievements

Achievement: Pair ExtraordinaireAchievement: Pull SharkAchievement: YOLO

Achievements

Achievement: Pair ExtraordinaireAchievement: Pull SharkAchievement: YOLO

Highlights

  • Pro

Block or report smadinen7

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.

Content in all repositories owned by your account will be closed.
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
smadinen7/README.md

Hi, I'm Sai Pranav Madineni 👋

AI Engineer · Data Scientist · CMU MISM-BIDA '25
Building eval-driven AI systems, multi-agent workflows, and financial ML pipelines.

About Me

  • 🎓 M.S. Information Systems Management (Business Intelligence & Data Analytics) — Carnegie Mellon University
  • 💼 AI Engineer Intern @ Virtual Gold Inc. — Built a production multi-agent GenAI financial assistant with eval-driven pipelines (+52% accuracy improvement)
  • 📦 Data Scientist Intern @ Amazon (Stores Finance) — Trend-break detection across 840+ P&L datasets; raised structural break precision from 50% → 78%
  • 🏦 Previously Software Engineer @ Societe Generale — Full-stack at one of the world's largest investment banks, cutting transaction processing time by 50%
  • 🎯 Focused on: LLM evaluation, RAG pipelines, financial time series, multi-agent systems

Tech Stack

AI & GenAI

PythonLangChainCrewAIHuggingFaceOpenAIPyTorch

Data Science & ML

Scikit-learnTensorFlowXGBoostPandasSQL

Cloud & MLOps

AWSDockerMLflowDatabricksAirflow

Featured Projects

ProjectWhat it doesStack
Financial AI Assistant — PrototypePublic prototype of a production multi-agent financial intelligence platform built at Virtual Gold Inc. Analyzes 10-K filings, generates investment recommendations, and benchmarks companies against peers.CrewAI · Gemini · Groq · Streamlit
Time Series TutorEval-driven RAG pipeline grounding LLM responses in time series research; 80% recall@5 via RAGASLangChain · FAISS · BM25 · Gemini
Prompt Engineering — FinanceCLEAR / Few-shot / Chain-of-Thought prompt design for financial analysis tasks across multiple LLMsPrompt Engineering · LLM Eval
DS · ML · AI Mastery14-phase, 100+ topic self-directed curriculum from foundations to production MLPython · PyTorch · MLOps

Experience Highlights

Amazon (Data Scientist Intern) → Trend-break detection on 840+ P&L datasets
Structural break precision: 50% → 78%
RMSE ↓ 10% | MAE ↓ 8% on cost forecasts
Virtual Gold Inc. (AI Engineer Intern) → Production multi-agent AI financial assistant
LangSmith + DeepEval observability pipeline
Model accuracy improved by 52%
Societe Generale (Software Engineer) → Transaction processing time cut by 50%
Full-stack: Java · Spring Boot · Oracle Cloud

Pinned Loading

  1. Prompt-Engineering-FinancePrompt-Engineering-FinancePublic

    CLEAR, Few-shot, and Chain-of-Thought prompt engineering for financial analysis tasks (information synthesis, Q&A) evaluated across multiple LLMs.

    1

  2. ceo-ai-assistantceo-ai-assistantPublic

    Public prototype of a production multi-agent AI financial intelligence platform. Analyzes 10-K filings, generates investment recommendations, and benchmarks companies using CrewAI + Gemini + Groq.

    Python

  3. ds-ml-ai-masteryds-ml-ai-masteryPublic

    14-phase, 100+ topic self-directed curriculum from Python/Math foundations to production MLOps, LLMs, AI Agents, and Reinforcement Learning.

    2

  4. forecasting-rag-learning-platformforecasting-rag-learning-platformPublic

    Eval-driven RAG pipeline for corporate forecasting and structural break detection. Achieves 80% recall@5 via RAGAS evaluation. Built with LangChain, FAISS, and BM25 reranking.

    Python 1