Skip to content
View JayacharanR's full-sized avatar
9 followers · 17 following
  • Vellore Institute Of Technology
  • Vellore, India
  • 01:49 (UTC +05:30)
  • LinkedIn in/jayacharan-r

Achievements

Achievement: Starstruck

Achievements

Achievement: Starstruck

Highlights

  • Pro

Block or report JayacharanR

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. 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
JayacharanR/README.md
Jayacharan R — Be kind. Stay curious. Retrieval, cloud, and world models. The JR atom sits above three yellow and gold traces on a shared baseline.
I build across AI, retrieval, and cloud infrastructure, with a soft spot for systems that can explain a decision, recover from a mistake, and hold up under scrutiny.

LinkedIn Email Explore my repositories

Selected builds

01 / Clean-Backup — Rust-powered perceptual deduplication, on-device AI, and reversible photo-library organization. Rust, Python, ONNX. Explore the repository. 02 / WikiMind — Wikipedia question answering with two-stage hybrid retrieval, live updates, reranking, and citation checks. Qdrant, Redis, LangGraph. Explore the repository. 03 / Carbon-Aware Cloud Optimization — Multi-region AKS scheduling with an LLM proposer, deterministic checks, and an optimization fallback. Azure AKS, Llama 3, PuLP, Python. Explore the repository. 04 / Auto Thinker Engine — Research in progress: camera-based driving in CARLA with JEPA representations, DreamerV3, and multi-camera navigation. PyTorch, JEPA, DreamerV3, CARLA. Explore the repository.
A closer look at the engineering ↴
Project The interesting part
Clean-Backup Rust/PyO3 hashing meets a Python pipeline. Local ONNX models handle classification; journaled file operations support session rollback.
WikiMind Discover articles, retrieve dense + sparse chunks, fuse and rerank, then check grounding and citations. Unverifiable answers become abstentions.
Carbon-Aware Cloud Compare proposed region/start choices against an exact feasible-candidate reference. Check inputs and constraints, then fall back or hold when needed.
Auto Thinker Engine Compare visual representations and world-model approaches under explicit driving gates. Research is paused; final driving qualification remains open.

Experience

GenAI Intern · Tata Consultancy Services (Jun–Jul 2026) — Worked on RAG search, prompt-injection guardrails, and Langfuse tracing. SAP Analytics Cloud Intern · Blueprint Technologies (May–Jul 2025) — Built SAP Analytics Cloud dashboards and predictive models. What connects the work. Useful intelligence needs good systems around it. That means fresh evidence for retrieval, explicit constraints for scheduling, reversible operations for personal data, and honest evaluation for learned policies.

✦ Tech stack ✦

Python, Rust, SQL, PyTorch, ONNX, OpenCV, Langfuse, SAP Analytics Cloud, Docker, Kubernetes, AWS, Azure, PostgreSQL, Redis, SQLite, Neo4j, Linux, Git, GitHub Actions, FFmpeg, LangChain, LangGraph, Hugging Face, Qdrant, Ollama, and CARLA

Certifications

AWS Certified Solutions Architect – Associate Microsoft Certified: Azure AI Engineer Associate Microsoft Certified: Azure Data Fundamentals

Contribution activity for the last year. Weeks appear in order as a bright particle traces days with activity.
Public GitHub contribution calendar · refreshed daily

Be kind. Stay curious.

Pinned Loading

  1. Clean-Backup Clean-Backup Public

    Intelligent media organizer with Rust-powered perceptual deduplication, on-device AI classification (scene + face recognition via ONNX), EXIF-based date organization, Google Drive sync, and a React…

    Python 19

  2. WikiMind-End-to-End-Hybrid-RAG-Pipeline WikiMind-End-to-End-Hybrid-RAG-Pipeline Public

    Wikipedia-grounded question answering with two-stage dense and sparse retrieval, Qdrant, Redis caching, LangGraph CRAG/Self-RAG orchestration, citation checks, evaluation, and a Streamlit dashboard.

    Python

  3. Carbon-Aware-Cloud-Optimization Carbon-Aware-Cloud-Optimization Public

    Carbon-aware benchmark scheduler for multi-region Azure AKS, combining Llama-3 proposals, deterministic trust gates, and exact optimization.

    Python

  4. Auto_Thinker_Engine Auto_Thinker_Engine Public

    Camera-based driving research in CARLA with JEPA encoders, DreamerV3 world models, and multi-camera navigation

    Python 1