AI Engineer with less than a year in LLM Engineering and Generative AI.
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
AI Engineer with hands-on experience building production-grade LLM evaluation systems, multi-agent pipelines, RAG infrastructure, and autonomous AI agents using Python, LangGraph, LangChain, LangSmith, OpenRouter, and Qdrant. Strong foundation in data analytics with proven ability to design evaluation frameworks, prompt engineering workflows, and real-time AI observability platforms.
Arya College of Engineering
B.Tech · Artificial Intelligence & Data Science
January 1, 2022 – June 1, 2026
ZIDIO Development
AI Engineer Intern
June 1, 2025 – August 1, 2025
null
Linux World Pvt. Ltd.
Generative AI & Machine Learning Intern
July 1, 2024 – August 1, 2024
Jaipur, Rajasthan, India
Multi-Agent Code Review & Auto-Remediation System
June 1, 2026 – Present
Orchestrated a 3-agent LangGraph pipeline (static analysis, OWASP security scanner, LLM fix proposal) with enforced per-review cost budgets under /bin/sh.50; autonomously identified 23 issues including 4 critical vulnerabilities (SQL injection, hardcoded secrets, insecure deserialization, shell injection) in benchmark testing. Implemented 12-pattern OWASP security scanner using regex + AST analysis (zero API cost), LangGraph typed state contracts eliminating silent LLM output failures, and structured fix proposals with diff-style patches for critical issues.
CodePulse - AI-Powered Cinematic Code Walkthrough
June 1, 2026 – Present
Built a full-stack AI web app sending code to LLaMA 4 via Groq, returning a structured ExecutionScript JSON powering a live line-by-line animated walkthrough with variable tracking, call stack, and voice narration; deployed on Vercel with Monaco Editor, timeline scrubber, and 4 ELI5→Senior explanation levels. Shipped "Break Mode" - AI-driven bug injection with red pulse failure animation and fix narration; animated SVG character narrators (Batman/Spiderman/Superman) with live mouth sync via Web Speech API; zero hallucination tolerance enforced via typed JSON + error recovery prompts.
LLM Evaluation Harness & Red-Teaming Framework
June 1, 2026 – Present
Engineered a production-grade LLM evaluation framework benchmarking GPT-4o and Claude Sonnet across 50 MMLU prompts using ROUGE-L, BERTScore, and Exact Match with bootstrapped 95% confidence intervals reducing model regression detection from days to under 4 minutes per CI run. Architected an LLM-as-judge ensemble pipeline (GPT-4o + Claude) with inter-rater agreement scoring (Cohen's κ = 0.81) and 20+ adversarial red-team attack patterns (prompt injection, jailbreaks, encoding bypasses), catching critical safety regressions before deployment; integrated MLflow experiment tracking and LangSmith prompt lineage tracing.
Real-Time RAG Ops Platform with Drift Detection
June 1, 2026 – Present
Built a production RAG pipeline with Qdrant vector search, deterministic n-gram embedding (384-dim, zero API cost), and cosine similarity drift detection triggering automated re-indexing when embedding distribution shifts beyond threshold; achieved 75% Redis cache hit rate on repeated queries. Deployed real-time ops dashboard (vanilla JS + Chart.js) tracking p95/p99 latency SLOs, retrieval quality (Hit@5), embedding drift score, and live event log demonstrating production observability skills beyond basic RAG tutorials.
Autonomous Financial Research Agent
June 1, 2026 – Present
Architected a tool-augmented autonomous AI agent executing a 4-step structured reasoning chain (Market Context → Financial Analysis → Risk Assessment → Investment Thesis) using Yahoo Finance and DuckDuckGo APIs with per-query LLM cost budget enforcement at /bin/sh.50. Implemented full reproducibility audit trail with tool call logging, SHA-256 reproducibility hash per report, and Redis caching reducing repeated query costs by ~60%; complete company research reports generated in under 60 seconds with structured JSON output.
Generative AI
Mastermind Outskill
June 1, 2026 – Present
Data Science & ML
Udemy
June 1, 2026 – Present
Data Analytics
June 1, 2026 – Present
Cultural Fit Analysis
The candidate's portfolio showcases a strong passion for AI engineering, with diverse projects spanning code analysis, financial research, and cinematic walkthroughs. This breadth of interest and proactive project development aligns well with an innovative and fast-paced AI engineering environment. The focus on practical, deployable solutions and rigorous evaluation methodologies suggests a results-oriented approach. The candidate's education in AI & Data Science further reinforces their foundational alignment with an AI-centric role.
Soft Skills & Operational Fit
The candidate's project descriptions highlight a strong problem-solving aptitude, attention to detail (e.g., cost budgets, reproducibility, drift detection), and an ability to work on complex, multi-faceted AI systems. The emphasis on production observability and robust evaluation frameworks suggests a mature operational mindset. The detailed project descriptions also indicate good written communication skills for technical concepts.