AI Product Manager

Hi, I'm Somraj.

I build AI-native products at scale — from LLM personalization engines to agentic systems — bridging deep technical execution with measurable business outcomes. 3+ years shipping at 10M+ user scale.

Scroll
Product StrategyNext.jsPythonSQLReactTypeScriptFigmaData AnalyticsAgileTableauPowerBIMachine LearningGenerative AIUser ResearchA/B TestingJiraProduct StrategyNext.jsPythonSQLReactTypeScriptFigmaData AnalyticsAgileTableauPowerBIMachine LearningGenerative AIUser ResearchA/B TestingJira

About Me

I'm an AI Product Manager who operates at the intersection of product strategy, machine learning systems, and data-driven growth. I've shipped LLM-powered personalization engines, owned full product lifecycles end-to-end — from discovery to deployment — and built agentic tooling that replaces entire manual workflows. My edge: I can write PRDs and API contracts in the morning, run multivariate experiments by afternoon, and review LLM eval frameworks by evening.

"AI is not a feature layer — it's a new product paradigm. The best AI products are built by PMs who can reason about model behavior, design feedback loops, and define what 'good' looks like for a non-deterministic system. I bring that lens to every roadmap I own."

Key Strengths

1
AI/LLM Product Strategy & Lifecycle
2
Agentic Systems & Workflow Automation
3
Experimentation at Scale (10M+ users)
4
Cross-Functional Technical Leadership
5
Data Storytelling & Funnel Forensics

Core Competencies

AI & LLM Product

LLM Integration & Evaluation (Evals)Prompt EngineeringRAG ArchitectureAgentic System DesignMCP (Model Context Protocol)AI Feature Lifecycle ManagementResponsible AI & Bias AwarenessOpenAI / Anthropic APIs

Product & Growth

0-to-1 Product StrategyOKR & Roadmap PlanningPRD & System Flow DocumentationAPI Contract DesignA/B & Multivariate TestingCRO & Funnel OptimizationFeature Prioritization (RICE/ICE)Agile / Sprint PlanningGo-to-Market Execution

Data & Analytics

GA4AmplitudeMetabaseMicrosoft ClaritySQL & Cohort AnalysisFunnel & Event TrackingPower BI / Looker StudioHeatmaps & Session Replays

Technical & Infrastructure

Next.jsGolangDocker & KubernetesAWS (EC2 / S3 / Lambda)GCP (200+ Labs)CI/CD & Feature FlagsREST APIs & WebhooksFirebase

Design & Collaboration

Figma (Wireframes & UX Audits)User Journey MappingJIRA & ConfluenceStakeholder CommunicationCross-functional Leadership

AI & Agentic Product Work

Hands-on shipping — not just strategy. From LLM pipelines to autonomous agents deployed in production.

CTMP AI Personalization Engine

Architected a survey-driven, LLM-assisted personalization system that maps user skin concerns to product routines at scale. Delivered +15% add-to-cart and -18% bounce rate across 10M+ monthly users.

AI-Powered Autonomous Audit Agent

Designed and deployed a headless browser agent using LLMs to autonomously crawl UI structures and generate heuristic UX audit reports — replacing 2-week manual processes with 48-hour turnarounds.

GPT-Integrated Project Tracker

Built a full-stack project management platform with GPT-powered task summarization and status workflows. Implemented JWT auth, API monitoring, and observability — production-ready from day one.

Featured Work

LLM · Rule Engine · Next.js · REST APIs

CTMP AI Personalization Engine

Problem

Users arrived at a D2C skincare platform with varied, complex skin concerns and received generic, non-personalized experiences — leading to high bounce rates and low add-to-cart.

Approach

Designed end-to-end architecture: a survey-driven concern intake layer, an LLM content generation pipeline for routine copy, and a rule-based product matching engine. Built for modular scalability so new concern categories can be added without re-architecture.

Impact

+15% add-to-cart rate, +11% PDP engagement, -18% bounce rate. Operating at 10M+ monthly user scale.

Figma · GA4 · Clarity · Agile · Core Web Vitals

Homepage & PLP Revamp

Problem

Homepage and PLP suffered from UX friction identified through heatmap analysis and Core Web Vitals data — causing measurable funnel drop-offs and slow page performance.

Approach

Owned full product lifecycle: discovery via heatmaps and funnel data, spec writing, UX redesign in Figma, front-end review, and multivariate testing across 3 agile sprints.

Impact

+7% conversion rate, +9% average order value, 22% faster page load.

Next.js · Golang · MongoDB · OpenAI API

AI Project Tracker

Problem

Teams lacked a unified task management system with intelligent summarization and built-in observability for distributed workflows.

Approach

Built a full-stack PM platform from scratch: GPT-powered task summarization, status workflows, JWT authentication, analytics dashboards, and API monitoring — leveraging OpenAI API for contextual task insights.

Impact

Production-deployed platform with full observability. Demonstrated end-to-end technical PM capability — from architecture to deployment.

How I Can Help

AI Product Strategy & Execution

End-to-end AI product work — from defining LLM evaluation frameworks and writing AI-specific PRDs, to shipping agentic workflows in production.

0-to-1 Product Incubation

Whitespace discovery to MVP. Translating raw market gaps into scalable system architectures with clear go-to-market paths.

Growth & Experimentation Systems

Building rigorous A/B and multivariate testing frameworks that systematically lift conversion, retention, and activation metrics at scale.

Professional Journey

Apr 2024 – Present

Senior Executive – Growth Product (AI Initiatives)

Honasa Consumer Ltd — Mamaearth & The Derma Co.

  • Architected CTMP AI personalization engine: LLM-assisted, survey-driven concern-to-routine mapping system — delivering +15% add-to-cart and -18% bounce rate across 10M+ monthly sessions.
  • Owned quarterly product roadmap using OKR methodology; prioritized features via funnel analytics, cohort analysis, and event-tracking data from Amplitude and GA4.
  • Led homepage & PLP revamp end-to-end — heatmap discovery, Figma specs, front-end review, multivariate testing — achieving +7% CR and +9% AOV across 3 agile sprints.
  • Governed platform reliability at flash-sale scale (10M+ sessions): CI/CD pipelines, feature flags, release checklists — 99.9% uptime, +25% deploy velocity.
  • Drove mobile-first strategy: redesigned onboarding flows, deep linking, and push notification logic — growing app session share from 34% to 48%.
  • Wrote detailed PRDs, system flow diagrams, and API contracts; led UAT cycles and sprint planning across engineering and design squads.
Jun 2023 – Apr 2024

Growth Product Intern

Honasa Consumer Ltd — Mamaearth & The Derma Co.

  • Stabilized checkout and order flows for 15,000+ concurrent users during high-traffic flash sales.
  • Ran 12+ A/B and multivariate experiments on UI flows; 4 shipped tests delivered sustained 3–5% CR uplift.
  • Reduced defect resolution time by 28% through structured bug triaging, test case creation, and cross-team release coordination.
  • Built behavioral analytics dashboards tracking NPS, funnel drop-offs, and KPIs using heatmaps, session replays, and SQL.

Let's Build Something Great

I'm currently engaged with new opportunities. If you have a question, a project in mind, or just want to discuss product strategy, let's connect.

© 2026 Somraj. All rights reserved.