Senior Executive — AI/ML & Generative AI · Nvest Solution

Aditya Naranje
builds production AI.

I'm an AI/ML Engineer with 2 years of experience designing, deploying and evaluating RAG systems, LLM agents and multi-agent workflows — with LangGraph, MCP tool calling and AWS — that ship to production and drive measurable automation.

View Projects Download CV
LangGraph · MCP RAG · pgvector
2+ yrs
Building production AI
85%
Less manual PR review
10+
RAG bots in production
9.14/10
B.E. CGPA
Portrait of Aditya Naranje, AI/ML Engineer Amravati, Maharashtra, India
About

Engineering AI systems that ship — and hold up in production.

I'm Aditya Naranje, an AI/ML Engineer with 2 years of experience designing, deploying and evaluating production-ready AI. At Nvest Solution I build LLM agents, multi-agent RAG services and MCP tool integrations for insurance workflows — with multi-tenant data scoping, human-in-the-loop review and rigorous model evaluation.

My toolkit spans LangGraph, LangChain, MCP, RAG, vector search (pgvector, Pinecone, FAISS), Python and AWS (Bedrock, Lambda, EC2, SageMaker), plus the ML layer — CatBoost/scikit-learn, Optuna tuning and champion/challenger promotion. I focus on translating business problems into scalable AI with measurable gains in automation, reliability and efficiency.

Production RAGMulti-tenant retrieval with hybrid BM25 + vector search, embeddings and evaluation.
LLM & multi-agent systemsLangGraph orchestration, tool calling, MCP servers and human-in-the-loop.
ML evaluation & monitoringAUC/F1-gated champion/challenger for churn, propensity and lead-scoring.
Cloud & CI/CDAWS deployment, Docker, GitHub Actions, FastAPI/Flask and serverless automations.
Generative AI LLM Agents RAG LangGraph MCP AWS Python
AI Engineering

What I build

The core capabilities behind my recent projects — end-to-end systems, not notebooks.

RAG Systems

Multi-tenant retrieval services grounding LLMs in private data — hybrid BM25 + vector search, embeddings and relevance evaluation.

Parse→ Embed→ Retrieve→ Generate→ Evaluate

// Multi-Bot RAG · Bedrock AgentCore · pgvector knowledge service

AI Agents

Stateful, tool-using agents orchestrated with LangGraph — including MCP servers — that plan, call tools and validate with human-in-the-loop review.

Plan→ MCP tools→ Execute→ Validate→ Review

// CodeGuard PR Agent · MCP onboarding server · Research Agent

LLM Applications

Conversational products — chatbots, content tools and copilots — with structured output and guardrails.

Prompting→ Structured output→ Guardrails→ Evaluate

// CODEVO · Fitness Chatbot · SparkLine · Story Generator

Workflow Automation

Event-driven automations connecting GitHub, Telegram, WhatsApp and email into intelligent notification flows.

Trigger→ Enrich (LLM)→ Route→ Notify

// GitHub Notifier · WhatsApp Agent · Telegram bots

Technical Expertise

Skills & tools

Organised by where I use them — from model to deployment.

AI & GenAI

llms, rag & agents

LLMs Generative AI LLM Agents RAG Multi-Agent Systems Prompt Engineering Embeddings Vector Search Tool Calling MCP Model Evaluation LangSmith NLP Transformers Hugging Face

Machine Learning & Data

models, tuning & eval

Machine Learning Deep Learning Regression Classification Scikit-learn CatBoost Optuna MLflow Weights & Biases Pandas NumPy Matplotlib Seaborn Data Visualization SQL

Agent Orchestration & Apps

frameworks & serving

LangGraph LangChain Langflow MCP Tool Calling Agent Orchestration Python FastAPI Flask Streamlit n8n

Cloud & Data Infrastructure

deploy & vector stores

AWS Bedrock AWS Lambda AWS EC2 SageMaker AgentCore Docker Git GitHub Actions CI/CD PostgreSQL / pgvector Pinecone FAISS Grafana Prometheus
Experience

Where I've worked

Currently building production AI in the insurance domain; earlier roles built my data and ML foundation.

Senior Executive — AI/ML & Generative AI

Nvest Solution Private Limited · Thane, MH
Jul 2024 – Present
  • Deployed an LLM-based agent (LangGraph, Flask) for automated GitHub PR review with severity scoring and inline fix suggestions — cutting manual review effort by ~85% and PR turnaround by ~70%.
  • Built a production RAG knowledge service (Docling parsing, Voyage AI embeddings, PostgreSQL/pgvector with hybrid BM25 + vector search) with multi-tenant data scoping across insurance clients.
  • Developed a local MCP server for insurance product onboarding with tool-calling/agent orchestration (Claude, Codex, Antigravity) — automating discovery, validation, diagnostics and SQL generation, cutting setup time ~95% (3–4 days → 2–3 hrs) with read-only DB access and human-in-the-loop review.
  • Built ML evaluation & monitoring (renewal-propensity, churn, lead-scoring) with scikit-learn/CatBoost, Optuna tuning and champion/challenger promotion gated on AUC/F1 — expanding recommendation coverage from 1 to 12 insurance lines of business.

ServiceNow Internship Trainee

Cognizant · Pune, MH
Dec 2023 – May 2024
  • Designed and implemented an approval workflow and automated notification system in ServiceNow — cutting leave-request processing time by 40% and eliminating manual follow-ups across the team.
  • Developed custom business rules, client scripts and workflow automations to streamline IT service operations, reducing average ticket resolution time and manual intervention.
  • Built and configured ServiceNow forms and workflows tailored to client requirements, improving process consistency and reducing data-entry errors across IT service pipelines.

Upskilling · AI/ML & Generative AI

Self-directed learning
Aug 2023 – Nov 2023

A focused period of self-directed learning between university and full-time work — strengthening AI/ML foundations and modern generative-AI, LLM and cloud skills through hands-on building.

Earlier experience · Data & ML internships (2021–2022)

  • Data Analysis Intern — iNeuron.ai Jan–Apr 2022
    Google Play Store data analysis; built a Python/Heroku dashboard web app.
  • Data Science Intern — LetsGrowMore Sep 2021
    Stock-price forecasting with LSTM; OpenCV image processing; Iris clustering.
  • Data Science Intern — iNeuron.ai Aug–Sep 2021
  • Computer Vision Intern — The Sparks Foundation Jul 2021
  • Machine Learning Intern — Internship Studio Jun 2021
  • Data Science & Business Analytics Intern — The Sparks Foundation Mar 2021
    EDA on IPL and terrorism datasets; cleaned & visualised sales data.
Projects

Featured work

Agentic and generative AI first, then machine learning and data analysis. Every project links to its source.

See all repositories on GitHub
Education

Academic background

Undergraduate · Graduated

B.E. Computer Science & Engineering

Prof. Ram Meghe Institute of Technology & Research, SGBAU · Amravati

2019 – 2023CGPA 9.14 / 10
Senior Secondary

Higher Secondary (XII)

Rajeshwar Union Junior College, Badnera

2018 – 201985.08%
Secondary

Secondary School (X)

Rajeshwar Union High School, Badnera

2016 – 201791.60%
Certifications

Credentials & training

Key certifications from Udemy, IBM, Internshala, LinkedIn Learning and Kaggle — plus verifiable certificate scans below.

Certificate scans

Writing

Latest on Medium

I write about RAG architecture, LLM engineering and machine learning to sharpen and share what I learn.

Read all articles on Medium
Contact

Let's build something intelligent.

I'm open to AI/ML engineering roles, collaborations and interesting problems. The fastest way to reach me is email or LinkedIn.

Email Me Download CV
aditya.naranje7@gmail.com LinkedIn GitHub Medium Kaggle