I'm an AI Engineer and Backend Developer focused on building intelligent software systems with Python, LLMs, RAG pipelines, agentic AI, and scalable backend architectures.

I enjoy working at the intersection of AI + software engineering—from evaluating LLMs and designing retrieval pipelines to building APIs, observability systems, and AI-powered developer tools.

AI Engineering
      +
Backend Systems
      +
LLMs / RAG / Agents
      =
Practical AI Applications

My goal is simple:

Build AI systems that are useful, reliable, observable, and engineered to work beyond the demo.

┌──────────────────────────────────────────────────────┐
│                 AI & LLM SYSTEMS                     │
│                                                      │
│  • Retrieval-Augmented Generation (RAG)              │
│  • Hybrid Search & Semantic Retrieval                │
│  • Agentic AI Workflows                              │
│  • LLM Evaluation & Validation                       │
│  • Prompt Engineering                                │
│  • Model Fine-Tuning                                 │
└──────────────────────────────────────────────────────┘

┌──────────────────────────────────────────────────────┐
│                 BACKEND & ENGINEERING                │
│                                                      │
│  • REST API Development                              │
│  • FastAPI                                           │
│  • Secure Request Validation                         │
│  • Structured Error Handling                         │
│  • Observability & Logging                           │
│  • API Testing & Reliability                         │
└──────────────────────────────────────────────────────┘

Jan 2026 – Jul 2026

• Developed and tested REST APIs with secure request validation, input sanitization, and structured error handling.
• Implemented structured logging and observability patterns to improve failure detection and reduce debugging time.
• Performed thorough API validation using Postman, covering edge cases and ensuring reliable data exchange.

Sep 2025 – Jan 2026

• Evaluated GPT, Claude, and Gemini across 150+ prompts using structured JSON-based validation frameworks.
• Identified 50+ edge cases affecting LLM response consistency and reliability.
• Built assertion-based checks to improve automated validation and model reliability.

Mar 2025 – May 2025

• Built SK Fruits, a Java Android application for inventory management and CRM.
• Developed a system used by 150+ users, including 140+ customers and 6+ administrators.
• Applied SOLID principles and MySQL-backed validation to improve maintainability and system integrity.

Understand unfamiliar codebases faster with hybrid retrieval and structural code intelligence.

Stack: Python FastAPI React FAISS BM25 Tree-sitter

• Built a Hybrid RAG pipeline combining dense and keyword retrieval for accurate codebase question answering.
• Implemented AST-aware chunking using Tree-sitter for more meaningful code context.
• Generated dependency graphs to improve contextual retrieval.
• Developed GitHub repository ingestion and indexing workflows.
• Indexed 300+ code chunks for context-aware code retrieval.

Turning logs and telemetry into actionable debugging intelligence.

Stack: Python FastAPI OpenTelemetry LLMs Streamlit

• Integrated OpenTelemetry observability with real-time monitoring dashboards.
• Built LLM-powered log analysis workflows for deployment failure investigation.
• Enabled proactive issue tracking through centralized observability.
• Reduced manual debugging effort by approximately 20%.

A multi-agent AI system designed to intelligently route patient support workflows.

Stack: Python LangChain RAG Groq LLMs FastAPI FAISS Sentence Transformers Streamlit

• Built a multi-agent assistant routing queries between intake and clinical agents.

• Implemented patient lookup across 25+ synthetic records.

• Developed RAG using FAISS + Sentence Transformers + Groq LLMs.

• Added a web-search fallback for queries outside the internal knowledge base.

• Built a FastAPI + Streamlit application with centralized logging of:

  • Agent decisions
  • Retrieval operations
  • Tool calls

Languages

Backend & Databases

AI & Machine Learning

vighnesh = {
    "role": "AI Engineer",
    "focus": [
        "LLMs",
        "RAG",
        "Agentic AI",
        "Backend Systems",
        "Observability"
    ],
    "languages": ["Python", "Java", "C++"],
    "mindset": "Build → Measure → Improve",
    "goal": "Turn AI ideas into reliable software"
}

🚀 Building reliable software systems and practical AI applications.