AI Engineer with 1+ years in Computer Vision, LLM, and Agentic RAG
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 1 year of experience building, training, and deploying production systems across Computer Vision, LLM fine-tuning (QLORA), and Agentic RAG. Proven track record delivering end-to-end ML solutions—spanning model fine-tuning, Agentic RAG and Multi-Agent Systems, and MLOps practices for reliable training, serving, and monitoring in production. Experienced in designing scalable microservices and AI APIs, with a focus on shipping robust, cloud-deployed systems that move from prototype to production.
Malnad College Of Engineering
Bachelor of Engineering · Computer Science
August 1, 2021 – June 30, 2025
Visukhi Innotech Pvt Ltd
AI Engineer Trainee
June 1, 2025 – Present
India
Silo Fortune Pvt Ltd
Software Development Intern
November 1, 2024 – May 1, 2025
India
Medical Coding LLM — ICD-10 Assignment
June 24, 2026 – Present
Developed a two-stage QLORA fine-tuning of Gemma 3 4B with 4-bit NF4 quantization (LoRA r=64 → r=128) via TRL SFTTrainer for ICD-10 assignment. Stage 1 achieved 96.4% token accuracy on ICD-10 ontology injection. Stage 2 added robustness to clinical variations. Deployed a live REST API on Vertex AI Endpoint at 99.98% confidence. (GitHub)
Codebase Intelligence Bot — Graph-Based Agentic RAG
June 24, 2026 – Present
Transformed a FastAPI codebase into a queryable knowledge graph (6,208 nodes, 26,938 relationships) via tree-sitter AST parsing. Implemented a LangGraph ReAct agent with 3 tools (Cypher, Qdrant, Tavily) and 3-tier memory. Achieved RAGAS metrics: faithfulness 0.83, recall 0.81, relevancy 0.87 (200 queries) with an average latency of 1.2s. (GitHub)
View ProjectAlgoMind - Production RAG System
June 24, 2026 – Present
Developed a production RAG system with structure-aware parent-child chunking and hybrid dense+sparse retrieval (Nomic 768-dim + SPLADE), RRF fusion, and cross-encoder reranking. Implemented a 4-layer Redis cache, reducing latency from 3,200ms to 1ms on hit. Achieved RAGAS metrics: faithfulness 0.86, recall 0.79, relevancy 0.84 (150 queries). Deployed on Google Cloud Run. (GitHub)
View ProjectGitHub PR Review Bot
June 24, 2026 – Present
Developed a webhook-triggered PR review bot with HMAC-SHA256 validation and a custom diff parser. Used LangChain LCEL (Groq Llama-3.3-70B) constrained to valid diff positions to eliminate hallucinated line numbers. Implemented Redis SHA-256 caching (3.8s → ~0s), enabling reviews in 30-60s. (GitHub)
Mathematics for ML & Data Science
Coursera, Serrano Academy
December 1, 2025 – Present
Machine Learning Specialization
Coursera, Stanford Online (Andrew Ng)
August 1, 2025 – Present
Development of AI-Powered Model for Oral Cancer Detection
IEEE
January 1, 2025 – Present
Python for Data Science
NPTEL Elite, IIT Madras
August 1, 2023 – Present
Cultural Fit Analysis
The candidate's portfolio showcases a diverse range of AI projects, from medical coding and disease detection to codebase intelligence and PR review bots, indicating adaptability and a broad interest in applying AI to different domains. The involvement in hackathons and publications further demonstrates initiative and a proactive learning attitude. The experience with both personal projects and an AI Engineer Trainee role suggests a strong alignment with an innovative and results-driven culture.
Soft Skills & Operational Fit
The candidate's project descriptions and experience highlight a strong focus on end-to-end solution delivery, performance optimization (e.g., caching, latency reduction), and MLOps practices. Collaboration within a 3-member team is mentioned, indicating some team-oriented experience. The detailed project descriptions suggest a methodical approach to problem-solving and a drive to achieve measurable results.