ZUBER
AI Product EngineerMumbai

ZUBER SHAIKH

Building AI systems that process millions of data points for real-world impact

Recognitions

OpenAI Buildathon — State FinalistOpenAI Buildathon — State Finalist
Claude Impact Lab — India's FirstClaude Impact Lab — India's First
UIDAI Hackathon — 1.2M RecordsUIDAI Hackathon — 1.2M Records
TechTeenz — 100+ Member ClubTechTeenz — 100+ Member Club
CodeThon CLI — 7 LLM ProvidersCodeThon CLI — 7 LLM Providers
OpenAI Buildathon — State FinalistOpenAI Buildathon — State Finalist
Claude Impact Lab — India's FirstClaude Impact Lab — India's First
UIDAI Hackathon — 1.2M RecordsUIDAI Hackathon — 1.2M Records
TechTeenz — 100+ Member ClubTechTeenz — 100+ Member Club
CodeThon CLI — 7 LLM ProvidersCodeThon CLI — 7 LLM Providers

Engineering Evidence

Measured impact across every project — not claims, receipts

1.2M+

Government Records Processed

Statistical forecasting across 36 states/UTs for UIDAI using ARIMA and Mann-Kendall trend tests

7+

LLM Providers Integrated

OpenAI, Anthropic, NVIDIA, DeepSeek, Together AI, Ollama, LM Studio — unified CLI interface

100+

Engineers & Students Impacted

Founded TechTeenz Technical Club — project-first, peer-led learning model scaled to 100+ members

Recognition

External validation of engineering capability

India's First Claude Impact Lab

June 2026Mumbai, India
Top 10% of applicants (200 of 2,000+)

Selected for India's inaugural Claude Impact Lab — an invite-only innovation program organized by Anthropic's Claude Community with the Kumbhathon Innovation Foundation and the Government of Maharashtra. Built alongside founders, AI researchers, and engineers from Reliance Jio, Upstox, J.P. Morgan, Aditya Birla Group, Ola, CRED, EY, KPMG, IITs, and BITS — designing AI solutions for public infrastructure challenges.

Kumbh Plus — an offline-first AI reunification platform for Kumbh Mela 2027.

OpenAI Academy × NxtWave Buildathon

2026Pune, India
State-Level Finalist

Advanced to the State-Level Round of a nationwide AI Buildathon — two-member team, 48 hours, from ideation to working prototype. Built Nyay Sahayak, an AI multilingual legal assistance platform tackling real-world problems for India under hackathon conditions.

Nyay Sahayak — AI-powered multilingual legal aid platform.

Projects

Detailed breakdowns of engineering decisions and outcomes

Nyay Sahayak

AI/ML

Nyay Sahayak

AI legal aid that turns hours of consultation into a single conversation — multilingual, sub-3s latency

PythonFastAPIReact
UIDAI Analytics

Data Science

UIDAI Analytics

1.2M Aadhaar records, 36 states, 12 seconds — production-grade statistical analysis for India's identity infrastructure

PythonStreamlitARIMA
TechTeenz Club

Leadership

TechTeenz Club

100+ members, zero budget. Founded a technical club around a project-first, peer-led model

CommunityEventsTech
Parishtha

Full Stack

Parishtha

Full-stack e-commerce for spiritual products — JWT auth, dynamic catalogues, 99.9% uptime

JavaScriptReactNode.js
CodeThon CLI

Developer Tools

CodeThon CLI

7 LLM providers, one terminal. Open-source CLI for multi-agent code generation

TypeScriptNode.jsAI Agents
Kumbh Plus

AI • Civic Tech

Kumbh Plus

Offline-first AI reunification for 80M+ pilgrims at Kumbh Mela 2027

ReactTypeScriptFastAPI

Philosophy

Principles that guide every system I build

Identity

Ships production systems, not prototypes.

Bridges AI capability and user need. Every project — from legal aid to reunification platforms — was built, deployed, and used by real people.

Identity

Turns ambiguity into architecture.

Given a vague problem and a deadline, builds clarity through prototyping, feedback loops, and iteration. Thrives where requirements are unwritten.

Identity

Measures everything that matters.

From latency benchmarks to engagement metrics, instruments every system. If it cannot be measured, it cannot be improved.

Identity

Builds for the 99%, not the 1%.

Shipped the 80% solution that changes how people work before polishing the edges. Optimization is layered in where it actually matters.

Identity

Leads through building, not titles.

Built a 100+ member technical club by writing code and running hackathons, not writing mission statements. Influence is earned through output.

01

Systems over heroes.

The most impactful code outlives its author. I build systems that are self-documenting, testable, and operable by people who have never met me. No bus-factor, no bottlenecks.

02

Simplicity is the ultimate sophistication.

Every feature request is a cost before it is a benefit. I push back on complexity by asking: does this make the system more valuable or just more complicated?

03

Data before dogma.

I have no religious attachment to frameworks, languages, or architectures. I choose tools based on what the data says about the problem, not what is trending on Hacker News.

04

Velocity is a feature.

A mediocre solution shipped today is worth more than a perfect solution shipped next quarter. Speed of iteration compounds — and every cycle teaches something new.

05

Quality is a property of the system, not the output.

You cannot test quality into a product after the fact. Quality must be designed into the architecture, chosen in the dependencies, and baked into every layer of the stack.

06

Build for the maintainer.

The person who will debug your code at 2 AM is future you. I write code that is kind to future readers: explicit over clever, consistent over creative, tested over trusted.

Experience

Problem → Action → Result

President

TechTeenz

Oct 2025 — PresentMumbai, India

Problem: Students at TSPDC had no hands-on technical community — no hackathons, no project culture, no bridge between classroom theory and real-world building.

Action: Founded TechTeenz Technical Club with a Project-First, Peer-Led learning model. Organized hackathons, bootcamps, and career readiness programs from scratch.

Result: Scaled from zero to 100+ active members in the first year, establishing a culture where students build real products instead of theoretical assignments.

  • Founded and scaled TechTeenz Technical Club at TSPDC, growing from concept to 100+ active members within the first year
  • Established a Project-First, Peer-Led learning model focused on building real products rather than theoretical assignments
  • Organized hands-on hackathons, technical bootcamps, and career readiness programs bridging academia and industry

Frontend Developer Intern

Humble Walking

Jun 2025 — Jul 2025Mumbai, India

Problem: The company needed faster, more responsive web pages and a reusable component library to reduce development time across client projects.

Action: Built 5+ production-grade responsive pages using React & Tailwind CSS. Created 10+ reusable UI components. Optimized frontend performance through code splitting and lazy loading.

Result: Improved load times by 25% and user engagement by 15%, while reducing development time by 20% across projects through the reusable component library.

  • Built 5+ production-grade responsive web pages using React & Tailwind CSS — delivering fast, accessible UIs for real-world users
  • Improved load times by 25% (Lighthouse CI: FCP from 4.2s to 3.15s) and user engagement by 15% (GA4 session duration, 2-week A/B test)
  • Developed 10+ reusable UI components, reducing development time by 20% (measured across 4 sprints via component reuse tracking)

Data Analysis Intern

Cognify Technologies

Jun 2025 — Jul 2025Mumbai, India

Problem: Marketing campaigns were not well-targeted, revenue forecasts were inaccurate, and the analytics team spent hours on manual data entry.

Action: Built regression models on historical sales data for revenue forecasting. Created automated ETL pipelines to eliminate manual data entry. Developed audience segmentation models for campaign targeting.

Result: Improved campaign targeting accuracy by 25%, revenue forecast accuracy by 20%, and saved 10+ hours per week through automated pipelines.

  • Improved marketing campaign targeting accuracy by 25% (conversion rate comparison, pre/post segmentation model rollout)
  • Increased revenue forecast accuracy by 20% (RMSE comparison: regression model vs previous quarter's manual forecasts)
  • Reduced manual data entry by 40% (tracked via time logs: 10+ hrs/week saved for the analytics team)

Let's Build Something

Open to AI engineering opportunities, collaborations, and technical discussions.