Kartik Puri — Technical Lead, Pune (IN) — 9 yrs shipping

Backend & AI systems, built to last.

I run engineering for Incentivate, an enterprise sales-commission platform — a team of six, four product suites, and nine years of Python, Flask, Redis and ClickHouse. The last two years have gone into agentic AI that actually runs in production, not demos.

I also ship products of my own — a live EV-charging map for India, a tool that measures how LLMs talk about brands, and an arena where autonomous AI agents compete.

Incentivate calculates sales commissions for enterprises in pharma, insurance, manufacturing and financial services. The hard part isn't the math — it's the orchestration: data pipelines, validations, approvals, audit trails, and reporting that has to hold up when every client's plan is different.

My approach: build the platform so our implementation team can set up each client through configuration instead of my team writing one-off code. SQL-driven dashboards rendered in sandboxed iframes (a pattern I proposed that became the standard dashboarding mechanism), YAML-defined data pipelines, a no-code plan builder.

I also push the engineering culture forward, deliberately and in order: static analysis first, then the CI to enforce it, then the tests. The timeline alongside is verifiable from git history, which is the way I like my claims.

2022
Enterprise SSO — SAML integrations for large-client onboarding.
  • Multi-tenant auth and session management across the platform.
  • Later rewrote the SAML processing layer in Go to escape Python xmlsec build pain.
2023
Introduced SonarQube — static analysis and quality gating.
  • Severe/critical findings tracked as first-class debt via GitHub issues.
  • Grew into an AI flow: finding → spec → fix → ready-to-review PR.
2023
Drove ClickHouse adoption for high-volume analytical reporting.
  • SQL-first pipelines replacing pandas-heavy processing; parquet + chdb followed.
  • Lazy pagination so dashboards handle 1M+ row queries without breaking a sweat.
2024
Early product AI — "Ask AI" and generated reporting.
  • AI report generation: structures, SQL and dashboard configs from a spec.
  • Cut report build time from multiple days to a few hours.
2025
Jenkins CI + automated tests, where there were none.
  • CI runs Sonar in the pipeline — closing the loop on the 2023 gating.
  • Designed a test rig to validate the unmodified engine against fixtures in containers, with SQL assertions.
2026
Agentic AI in production — agents, not demos.
  • Unified Agno-based agent framework: NL→SQL agents, persistent sessions, per-agent model config.
  • In-product AI assistant with an analytics + feedback loop; an agentic playground for builders.
  • Redis pub/sub fan-out that promotes config across load-balanced servers with per-server acks.

hover an entry for the detail

5–6engineers led
4product suites
9 yrsshipping software
2 yrsagentic AI in production
02

Side projects

Shipped for real, not prototypes

Plus a Product Hunt-launched publishing tool, a UI library, a Chrome extension and more — all projects →

03

Writing

Engineering notes from real builds

All writing →

04

Stack

Honest depth, not a keyword wall

Backend — the core

  • Python — Flask, SQLAlchemy, MongoEngine
  • Redis + RQ — queues, workers, schedulers
  • REST APIs, auth/SSO, long-running job orchestration
  • MySQL, PostgreSQL, MongoDB
  • ClickHouse — OLAP, SQL-first pipelines, chdb, parquet

AI systems

  • Agentic workflows in production — Agno, LangChain
  • AWS Bedrock, Gemini, OpenRouter
  • NL→SQL agents, AI report generation, RAG
  • MCP, prompt systems, eval loops

Frontend & infra

  • Vue 3 / Quasar — strongest; React, Tailwind
  • Dashboards, builder UIs, data-heavy tables
  • AWS EC2/S3, Lambda (serverless ETL), Docker, Nginx, PM2
  • Jenkins CI, SonarQube, GitHub flow
05

Path

B.E. Computer Science, VIIT Pune, 2017
2023 — present Technical Lead — Backend & AI Systems · Aurochs Software / Incentivate Lead a team of 5–6 engineers across four product suites of an enterprise incentive-compensation platform. Own architecture and roadmap with the CTO; drive the platform's AI direction.
2020 — 2023 Solutions Consultant · Incentius Client-facing full-stack work across healthcare, pharma, fintech and eCommerce. Led a Web3 NFT marketplace build (Flask + Quasar + Solidity) with a small team.
2017 — 2020 Solutions Developer · Incentius Search systems (Elasticsearch), BI dashboards, data pipelines, mapping tools. Python, Vue, MySQL.
2014 — 2017 Co-Founder · Mystery Monks Early tech and operations during the startup phase, alongside college.

Off the keyboard: long-distance motorcycle touring and strategy board games — the same plan-and-optimize itch, different terrain.