Nandu Vamsi Krishna Yanamandala
Open to work1 years experienceData Scientist
Data Scientist specializing in AI, LLM development, and full-stack engineering.
Standing Rank
Rank Not Available
Developer Badges
Skills & Technologies
44 skillsWork Experience
Data Scientist
Solis Technology Pvt Ltd
- End-to-end SaaS platform digitizing the insurance distribution lifecycle - lead capture, customer KYC, quote generation (Motor/Health/Life/Commercial), policy issuance, commission settlements, and analytics across role-specific apps for Agents, RMs, Ops, HR, and Admin.
- Architected Nx monorepo with 13 NestJS microservices and 5 Next.js apps; event-driven design with gRPC inter-service calls, Temporal workflows, Redis pub/sub, and SSE/Socket.IO for real-time updates.
- End-to-end AI-assisted development using Claude Code across 100k+ LOC - architecture, implementation, debugging, and refactoring; integrated multi-tenant auth, KYC, quoting engine, settlements, and email/SMS/WhatsApp notifications.
- Designed an AI-powered no-code platform to automatically convert Figma screens into production-ready full-stack code using an agent-based architecture with MCP and LangChain for UI component and layout analysis.
- Integrated multiple LLMs (Mistral 7B, Gemini 2.5 Pro via Ollama) with RAG pipelines to improve component reuse and design-to-code fidelity, generating optimized React/Next.js components with responsive styling.
- Built an automated document masking solution using Tesseract OCR + regex pattern detection to identify and mask sensitive Aadhaar/PAN fields from scanned images and PDFs; applied OpenCV for precise region masking while preserving readability, with AWS S3 for secure storage.
- Extended pipeline into a SmartDoc AI Parserusing FastAPI, AWS Textract, and Google Gemini API to extract structured PAN/Aadhaar fields (Name, DOB, PAN) and export to CSV via Pandas.
- Built a multilingual voice-enabled chatbot supporting Kannada, Punjabi, Hindi, English, Telugu, and Tamil with real-time speech recognition; implemented ASR pipeline using NeMo Conformer with NLU models and TensorFlow/Keras for intent classification.
- Built a cross-platform mobile app with React Native automating customer declarations via live video sessions, real-time AWS Transcribe transcription, and ML-based data validation against customer records using Python FastAPI microservices.
Data Science Intern
Cognifyz Technologies Pvt Ltd
- Feature Analysis: Engineered and analyzed key features impacting restaurant ratings across 10,000+ entries; performed statistical profiling of categorical variables to identify distribution patterns and data quality issues for downstream modeling.
- Geospatial ML: Applied GIS-based feature extraction, hot spot analysis, and spatial autocorrelation to quantify location-rating correlations across cities and countries, generating location-aware features for predictive modeling.
- Predictive Modeling: Designed, trained, and validated a regression model to predict aggregate restaurant ratings from engineered features, achieving 80% accuracy via iterative hyperparameter tuning and k-fold cross-validation.
Projects
Hybrid Telegram GenAI Bot (RAG + Vision)
- Built a fully local GenAI Telegram bot combining a RAG pipeline (Mistral 7B + sentence-transformers + SQLite vector store) with multimodal vision (LLaVA 7B) for image captioning and keyword tagging - all running locally via Ollama with no cloud API dependency.
- Implemented query caching, top-k semantic retrieval with source attribution, and per-user message history for context-aware answers; structured as a clean modular pipeline (ingest → embed → retrieve → generate)
HireHive
- Dynamic job-scraping pipeline using Selenium and LinkedIn APIs; integrated Google Gemini API for automated skill extraction.
- Fine-tuned ML models with TF-IDF and N-grams for resume analysis; deployed as a Flask web app.
Math Agentic RAG
- Built a Human-in-the-Loop agentic RAG system using LangChain, HuggingFace, and Qdrant to solve math problems from datasets like GSM8K and MathQA. Integrated Flan-T5 with DSPy for prompt chaining and semantic retrieval.
- Enabled web search fallback via Exa API for out-of-domain queries and added CLI-based feedback loops for iterative learning. Achieved scalable, intelligent reasoning with structured retrieval and LLM-based generation.
DrowzeAI
- Real-time driver monitoring system combining computer vision (Dlib EAR/MAR, DeepFace) and audio analysis (SER + MFCC) for drowsiness and emotion detection; deployed multithreaded voice/GUI alerts via pyttsx3 and Tkinter.
Leaderboard Standings
Leaderboard Position Pending
Global test scores, peer standing percentiles, and algorithm leaderboard ranks are updated dynamically.
Assessment Highlights
Assessments Not Completed
Coding evaluations, system assessment results, and conceptual score badges will appear here after taking a test.
AI Collaboration Score
AI Collaboration Score Pending
Developer coding behavior, assistant cooperation, and AI pair-programming indicators are evaluated during live coding sessions.
Role Compatibility Profile
Role Compatibility Analysis Pending
Custom matching reports, candidate role compatibility percentiles, and core engineer strength profiles are processed once conceptual code screenings are complete.
Achievements
Achievements Not Earned
Special rewards, developer badges, system recognition certificates, and conceptual milestones will display here.
About Details
Professional Bio
Data Scientist with expertise in AI-powered SaaS platforms, LLM development, and computer vision. Experienced in architecting microservices, building RAG pipelines, and deploying scalable machine learning models. Passionate about end-to-end AI-assisted development and intelligent automation.
Bachelor of Technology in Computer Science in Computer Science
Gurukul Kangri Vishwavidyalaya (2021 - 2025)
Intermediate Board in Intermediate
SriGayatri Junior College (2019 - 2021)