Kartik Kumar
Available for freelance

AI automation
engineer.

I build bots, workflows, and CRM / e-commerce integrations that cut manual work — end-to-end, from logic to deployment.

AI/ML Associate at QHT Clinic (Regrow Clinic LLP) · Uttar Pradesh, India.

Kartik Kumar — AI Automation Engineer

Experience

Where I’ve built.

Two years of shipping AI automation across ops, e-commerce, and internal tooling.

Oct 2025 — Present

AI/ML Associate

QHT Clinic (Regrow Clinic LLP)

  • Built reorder reminders, order follow-ups, calendar event managers, and lead distribution pipelines — webhook-driven, scheduled, with proper error handling.
  • Developed and trained WhatsApp and Telegram bots using prompt engineering, RAG chains, and image OCR (DeepSeek), integrated with Claude and OpenAI.
  • Wired Supabase vector stores, MongoDB embeddings, and LeadSquared CRM — lead creation, updates, and all operations through APIs on the backend.
  • Shipped Shopify automations: order tracking, abandoned cart recovery, and reorder reminders over WhatsApp / email, with live data flowing to Sheets.
  • Python
  • n8n
  • Claude
  • OpenAI
  • DeepSeek
  • Supabase
  • MongoDB
  • LeadSquared
  • Shopify
  • WhatsApp API

Feb 2026 — Mar 2026

Freelance AI/ML Engineer

Independent

  • Built a central AI brain via Anthropic MCP — connected Sheets, CRM, files, and reports to Claude so the owner could query daily business performance in natural language. Cut manual status-check time by about 70%.
  • Fine-tuned a deep-learning fraud detection model with careful feature engineering and hyperparameter tuning — achieved 93% precision on anomaly flags while reducing training time by 40%.
  • Deployed an OpenClaw-based automation stack on a Hostinger Linux VPS with a Telegram chatbot for file, employee data, and key document management — replacing repetitive admin work and manual handovers.
  • Anthropic MCP
  • Claude
  • Python
  • Scikit-learn
  • Hostinger VPS
  • Linux
  • Docker
  • Telegram Bot

Aug 2024 — Nov 2024

AI Research & Automation Intern

Drop Organization

Remote

  • Conducted in-depth research on AI-powered coding assistants like GitHub Copilot and Bolt.new.
  • Developed automation scripts to analyze the impact of AI tools on software development efficiency.
  • Worked with REST APIs to integrate AI-based debugging features into web applications.
  • Python
  • REST APIs
  • GitHub Copilot
  • Bolt.new

Toolkit

Skills & tools.

Programming & Scripting

  • Python (Pandas, NumPy, Scikit-learn)
  • Java
  • JavaScript / TypeScript
  • Shell scripting

Machine Learning & Data

  • Supervised / unsupervised learning
  • Feature engineering
  • Model evaluation (precision, recall, ROC-AUC, F1)
  • Hyperparameter tuning
  • Anomaly detection
  • Time-series basics

LLMs & Generative AI

  • Prompt engineering
  • RAG pipelines
  • Tool-calling / agents
  • OpenAI
  • Claude
  • DeepSeek
  • LangChain-style workflows

AI Automation & Orchestration

  • n8n
  • Zapier / Make-style logic
  • Webhook design
  • Retry & error-handling patterns
  • Scheduled jobs
  • Multi-step workflow orchestration

Databases & Storage

  • Supabase (Postgres, vector stores)
  • MongoDB
  • Firebase / Firestore
  • Basic SQL
  • Redis caching

Cloud & DevOps

  • AWS EC2
  • Basic Linux administration
  • VPS deployment (Hostinger)
  • Docker fundamentals
  • SSH
  • Nginx reverse proxy

Messaging, CRMs & E-commerce

  • Interakt
  • LeadSquared
  • WhatsApp Business API
  • Telegram bots
  • Shopify
  • Payment / checkout webhooks

About

A little about me.

Currently

AI/ML Associate

QHT Clinic · Since Oct 2025

Location

Uttar Pradesh, India

Open to remote work

Education

Bachelor of Computer Applications

MJP Rohilkhand University, Bareilly

2022 — 2025

Board of High School & Intermediate Education

Uttar Pradesh, Prayagraj

2020 — 2022

I’m an AI automation engineer with hands-on experience building bots, workflows, and CRM / e-commerce integrations that cut manual work and move business outcomes. Most of my work sits between LLM tooling, workflow engines, and the messaging fabric that ties them together.

I take projects from scratch to full deployment using Claude Code or Codex, and I spend most of my curiosity budget experimenting with new AI platforms to figure out what actually works in production. If a manual process is eating hours, I want to build the system that replaces it.

Recent outcomes

  • Cut manual status-check time by about 70% for a small business owner using an Anthropic MCP-based AI brain that connected Sheets, CRM, files, and reports to Claude.
  • Achieved 93% precision on fraud detection anomaly flags with a fine-tuned deep-learning model.
  • Reduced model training time by 40% via careful feature engineering and hyperparameter tuning.

FAQ

Common questions.

What does an AI Automation Engineer do?

An AI Automation Engineer designs, builds, and deploys production systems that combine LLMs (like Claude and OpenAI), workflow engines (n8n, Zapier), and messaging channels (WhatsApp, Telegram) to automate business processes end-to-end — from customer intake and CRM sync to e-commerce order flows and internal reporting.

What tools does Kartik Kumar specialize in?

Kartik works day-to-day with Python, Claude and Anthropic MCP, n8n workflow automation, Supabase (Postgres and vector stores), MongoDB, DeepSeek OCR, WhatsApp Business API, Telegram Bot API, LeadSquared CRM, and Shopify. The full toolkit is listed in the Skills section on this page.

Is Kartik available for freelance AI automation work?

Yes. Kartik is currently AI/ML Associate at QHT Clinic (Regrow Clinic LLP) in India and takes on selective freelance builds around agentic workflows, RAG systems, and Linux-hosted ops stacks. Typical reply time is within a day.

Contact

Let’s build something.

Have a workflow to automate or an agent to ship? Send a note — I read everything, and I usually reply within a day.

Based in Uttar Pradesh, India· Open to remote work worldwide.