SDE-2 at Oracle → Product Manager

I’ve spent years building the “how.” Now I’m focused on the “what” and “why.”

Software Engineer with 5+ years of experience building AI, cloud, and enterprise systems — now transitioning into Product Management to work closer to users, problems, and product decisions.

See the work Get in touch

CurrentlyOracle

5+ yrs

Writing code, running discovery, and slowly convincing people I can do both.

Currently

SDE-2 · Oracle

Previously

IBM

Domains

AI · Cloud · Enterprise

Status

Open to PM

The bridge

I spent years making systems fast and reliable. Now I spend my time making sure we’re building the right thing before we build it right. Both skills come in handy.

B.Tech CSE, KIIT BhubaneswarOCI 2024 Generative AI Certified Professional

Engineering roots

Python · FastAPIRAG · LlamaIndex · LangChainGenAI · Prompt EngineeringOAuth 2.0 · IDCSDocker · Jenkins · CI/CDOpenTelemetry

Product lens

RoadmappingDiscoveryPricing & packagingStakeholder alignment

Real product experience

Where engineering met product

Over 5+ years at Oracle and IBM, I’ve worked on problems involving enterprise, AI scalability, performance, security, and production readiness.

These experiences taught me to look beyond implementation and think about user experience, business scale, trade-offs, dependencies, and measurable outcomes.

Experience 01 · OracleEnterprise onboarding

Scaling Enterprise Onboarding

4,000+

enterprise customer tenants

The problem

Associating application resources with every customer could have created 12,000+ backend associations—making each new customer add cost and complexity.

Why it mattered

Enterprise growth needed a path that preserved secure tenant boundaries without duplicating the product’s infrastructure.

Key insight

Customers needed isolated identity and access—not separate copies of the application.

My role

I worked with the platform team to align authentication, routing, and security around a shared global backend and UI.

Customer tenants

Unique identity

Secure routing

Shared access layer

Application

Global backend + UI

4,000+

Tenants supported at enterprise scale

12,000+

Resource associations potentially avoided

Shared model

Customer growth decoupled from resource growth

Key takeaway

Scaling a product does not always mean adding more infrastructure. Sometimes the real problem is redesigning the model so growth no longer multiplies complexity.

Experience 02 · OracleAI experience at scale

Scaling the AI Response Experience

The AI response flow struggled below 100 concurrent users. As traffic rose, slower responses and higher failure rates put the customer experience at risk.

10×

concurrency

From struggling below 100 to supporting 1,000 concurrent users

40% → 0.6%

failure rate

Reduced request failures under load

~6 sec

faster

Cut latency from the AI response path

What I found

01Blocking work caused concurrent requests to wait.
02Expensive resources were recreated during requests.
03Shared AI capacity made latency and rate limits unpredictable.

The decision

I treated the response lifecycle as one experience rather than three isolated issues: improve concurrency, reuse costly resources, and move latency-sensitive workloads to dedicated AI infrastructure.

The goal was not simply a faster backend—it was a predictable AI experience as usage grew.

See the technical decisions behind the result

Cooperative concurrency

Allowed waiting workloads to yield rather than serializing execution.

Resource reuse

Cached clients, vector stores, metadata, and LLM resources across the response lifecycle.

Dedicated capacity

Reduced rate-limit pressure for inference and embedding workloads.

Cross-functional execution

Taking the product from “working” to production-ready

4+ teams aligned

Performance Testing

Load tests, bottleneck discovery, and validation under higher traffic.

Security / PEN Testing

Enterprise security assessment and production-readiness standards.

Platform Engineering

Authentication, onboarding, routing, and infrastructure capabilities.

SRE

Deployment readiness, observability, operational checks, and reliability.

My role increasingly extended beyond implementation into coordinating the dependencies needed to validate performance, security, reliability, platform compatibility, and observability together.

Engineering taught me how to build systems.
These experiences taught me how to think about products.

What problem are we solving?

What prevents this from scaling?

Which trade-offs should we make?

How do we know it worked?

That shift—from asking only “How should we build this?” to considering the problem, trade-offs, people, and outcomes—is what led me toward Product Management.

Self-driven work

Personal case studies

3 case studies · full decks included

Capstone · Product Strategy2025

Zepto Cafe — From Trial to Habit

Built a repeat-confidence strategy for office snack-break users: Snack Mode organized around snack windows, freshness & trust labels, and one-tap reorder — all inside the existing Zepto Cafe experience. Backed by 30 user interviews, a competitor teardown, and survey data.

12% → 25% 30-day repeat rate≈ ₹7 cr/yr added GMV at 100k users

Read the full deck User survey Live prototype

Product Strategy2025

Zomato product strategy case study cover

Zomato — Reimagining User Engagement

A strategy to deepen retention and build a healthier relationship with food: comfort vs. discovery modes, home-cooked meal delivery, ingredient-level customization, and transparent pricing for price-conscious users. Grounded in market research and user research.

2 modes Comfort & Discovery UXNorth Star successful orders delivered

Read the full deck

UX Research · Product Strategy2025

Uber — Solving the Last 50 Meters

Diagnosed pickup friction in complex, high-density environments (airports, malls, events) and proposed smart pickup-point recommendations, zone-based pickups, last-50m navigation, and predictive nudges. Based on user survey and market analysis.

50% Uber cab market share analyzed$23.98B India taxi TAM (2026)

Read the full deck User survey

Research, not guesswork

Every number has a source.

These case studies are built on user interviews, SurveyMonkey responses, competitor teardowns, and market sizing — not back-of-the-envelope guesses.

Zepto survey ↗Uber survey ↗

30+

user interviews

SurveyMonkey studies

competitors teardown

live prototype

Let’s talk

Building something worth talking about?

I’m open to product roles where technical fluency and product judgment matter.

chetan.anand710@gmail.com Download résumé LinkedIn ↗