Kartik · Senior Engineer · AI Systemskrtk.dev / 2026 · ed.01

Kartik.

I design and ship AI-powered systems used by enterprise teams — from architecture to deployment.

01 — Experience
8+ years in production
02 — Currently
Tech Lead at Incentivate
03 — Building
ChargeDisha · LLMNarrative · Molt
04 — Available for
Senior eng & AI roles
edition · 01
krtk · 2026
§ 01 / Now

In the build, today.

Three products I'm shipping right now. ChargeDisha for honest EV charger data; LLMNarrative + Molt for honest AI evaluation.
Closed alpha · Live

ChargeDisha

// Community-verified EV charging map · India

A charging map kept honest by the last driver who actually plugged in — not by an API ping. Photo-gated check-ins, EXIF + GPS validation, anti-fraud built into the data model. Closed testing on Google Play; public launch imminent.

FlutterFirebaseCloud RunOla MapsPython
Request access →
fig.03 · highway uptime · pune ↔ mumbai
Lonavla corridor
↑ 47 verified · 7d
PuneTalegaonLonavlaKhandalaMumbai
60kW30kW60kW50kW30kW
01Lonavla DC fast60kW✓ live3m
02Khandala HW50kW✗ dead28d
03Mumbai-Pune E130kW✓ live1h
04Talegaon Ola60kW⚠ slow2h
$verify ./session.jpg · EXIF + GPS+pts
fig.01 · brand mention coverage
acme.co — weekly
↑ 191% · 12-wk
GPT-4o
82%+12
Claude
64%+8
Gemini
41%−3
Llama-3
28%+5
Mistral
19%0
Building · Live

LLMNarrative

// AI Narrative Intelligence Platform

A system that evaluates how language models reference your brand, product, and industry context across multiple LLMs. Surfaces strengths and risks in generative model outputs, weekly, per tenant.

PythonOpenAIAnthropicPostgresNext.js
Read the case study →
Building · Live

Molt Olympics

// Agent-Native Evaluation Framework

A daily challenge arena where autonomous agents compete via API submissions, structured identity, and human-in-the-loop scoring. Most AI systems lack transparent, repeatable evaluation. This builds it.

FastAPIPostgresRedisReact
Read the case study →
fig.02 · live leaderboard · day 142
Refactor & reason
847 submissions
01agent.molt· YOU1492+18
02claude-runner-31471+9
03gpt-coder-v21438−3
04llama-tinker1402+5
05swe-bench-bot1389−1
$molt submit ./agent.tar.gzSKILL.md
§ 02 / Archive

Selected work, shipped.

Side products and client systems. From Web3 publishing to Chrome privacy tooling — same pattern: identify the leak, build the bridge.
[ 01 ]
BlogtoNFT
// Web3 publishing platform · ProductHunt Top 10

Converts blog posts into mintable NFTs — bridging content ownership and blockchain distribution.

2022read →[ 02 ]
Tales of Dreams
// AI storytelling engine · async

Generates personalized bedtime stories using dynamic character injection and audio synthesis.

2023read →[ 03 ]
Barley UI
// Zero-dependency UI library · ~8KB

Semantic UI library for performance-first interfaces. No framework overhead, clean markup.

2023read →[ 04 ]
Click to Blur
// Chrome extension · privacy

Prevents revenue leaks by masking sensitive Stripe & Supabase data in public demos.

2024read →
§ 03 / Writing

Notes from the build.

Patterns I find worth documenting — performance, AI workflow, data engineering. Long-form, no fluff.
2025 · 03 · 21AI

How Repomix Made My AI Coding Workflow So Much Easier

Repomix: The Tool That Helps Me Talk to AI Better Lately, I’ve been using a tool called Repomix a lot—and it’s quickly become a regular part of how I work. If you ever try to explain your codebase to an AI assistant like ChatGPT or Claude, you know h...

3 min · essay →
2024 · 02 · 14ClickHouse

Using clickhouse to transform tick data into OHLC

Introduction In the world of finance, tick data represents the heartbeat of the market, capturing every price change and transaction that occurs. This granular data is invaluable for high-frequency trading strategies, market analysis, and understandi...

3 min · essay →
2023 · 11 · 03Python

Boosting JSON Performance in Python: A Deep Dive into orjson

Introduction Background on JSON in Python JSON, short for JavaScript Object Notation, has become a ubiquitous data interchange format, owing to its simplicity and lightweight nature. In Python, the standard library provides a module aptly named json ...

9 min · essay →
2023 · 09 · 21Python

Python Functions: Writing Better Code with Ease

Introduction Programming is all about functions. They help you organize and reuse code. Python is a popular programming language that has a lot of tools for working with functions. In this guide, we're going to explore Python functions from the basic...

5 min · essay →
§ 04 / Experience

Eight years, four chapters.

Tech Lead today; co-founder, consultant, and developer in the chapters before. Each step compounded into the system architect I am now.
Sep 2023 – PresentAurochs / Incentivate Solutions

Tech Lead Incentivate

Led engineering for enterprise incentive management platform.

  • ›Headed 10-developer team
  • ›Defined technical roadmap with CTO
  • ›Implemented structured PR + release discipline
  • ›Architected AI copilots for analytics & reporting
  • ›Managed deployment pipelines on EC2
2020 – Sep 2023Incentius

Solutions Consultant

Built an end-to-end Web3 application from scratch, leading a small team and owning the full-stack architecture.

  • ›Led a team of 2 developers to build a complete Web3 application—frontend, backend, smart contract integration, and deployment.
  • ›Designed the entire system architecture using Vue.js for the UI and Flask + MySQL for backend APIs, authentication, and metadata services.
  • ›Integrated Solidity smart contracts to enable NFT minting, ownership logic, and blockchain interactions.
  • ›Built a secure admin interface for minting tokens and managing metadata.
  • ›Implemented seamless wallet connection flows, transaction handling, and blockchain event tracking.
2017 – 2020Incentius

Solutions Developer

A strong foundation phase where I built search systems, dashboards, data pipelines, and full-stack web apps.

  • ›Developed complex backend + frontend applications using Python, Vue.js, AngularJS, MySQL, and Elasticsearch.
  • ›Designed and implemented data cataloging and profiling systems — including ingestion pipelines, log dashboards, and search-heavy interfaces.
  • ›Built BI dashboards that leveraged Elasticsearch logs and relational data to visualize KPIs and insights.
  • ›Created admin dashboards and automated scripts using Python + Pandas for daily data processing.
Jun 2014 – Jun 2017Mystery Monks

Co Founder

Handled early tech and operations during the startup phase.
§ 05 / Toolkit

What I reach for.

The set I'm productive with. Stable choices for production systems, with willingness to add new tools when the problem warrants it.
Architecture
  • System design
  • Async workers
  • Event flows
  • Data pipelines
AI & Data
  • OpenAI API
  • LangChain
  • RAG
  • ClickHouse
  • ElasticSearch
  • Pandas
Backend
  • Python
  • Flask
  • FastAPI
  • Redis
  • RQ
  • SQL
Frontend
  • Vue
  • Quasar
  • React
  • Tailwind
Infra
  • Docker
  • EC2
  • Nginx
  • CI/CD
  • Git
§ 06 / Contact

Building AI systems that ship — not demos.

If you're hiring for a senior engineering or AI product role, I'd like to hear about it.

→ LinkedIn / in/kartikchop→ GitHub / kartikpuri95
© 2026 Kartik · krtk.devedition 01 · designed with precision