Pranav DeoStaff Product Manager at Rakuten

The world behind the work

Brands seek growth.
People seek relevance.

I connect them through
the products I build.

Home
Résumé (PDF, opens in a new tab)

Product leadership / Sunnyvale, California

Pranav Deo.

Staff Product Manager at Rakuten

I turn advertiser goals into products that work for customers.

For more than five years at Rakuten, I have worked across the full product journey: understanding advertiser needs, launching Programmatic Loyalty, and shaping the recommendations and systems behind the experience. My work now also includes shared decisioning and AI-assisted creative workflows.

Explore my workRésumé (PDF), opens in a new tab

The product question

A business wants growth.
A customer wants something worth choosing.

The product has to make both possible.

  1. 01Understand the goal

    Advertiser intent and customer need

  2. 02Design the decision

    Recommendations, incentives, and controls

  3. 03Measure the outcome

    Business value and customer response

See it in a real product

Computer science trained me to understand systems.
Statistics trained me to question the evidence.

My background

Selected professional work

From a commercial problem
to a product people can use.

Two views into my work at Rakuten: a product I launched and the platform scope I lead today.

Advertising & marketplaces

Programmatic Loyalty: from advertiser goals to customer rewards.

I took Programmatic Loyalty from idea to launch, connecting advertiser budgets and return targets with commission recommendations, campaign pacing, and customer rewards.

My ownershipProblem discovery, MVP definition, advertiser workflow, cross-functional alignment, and phased delivery.

Read the case study: Programmatic Loyalty: from advertiser goals to customer rewards.Senior Product Manager
$500M+

Annual product revenue

Programmatic Loyalty, the product I launched
  • Advertiser experience
  • Recommendations
  • Marketplace economics

AI & platform leadership

Making shared decisions useful. Making AI workflows reviewable.

My current scope brings together shared recommendation systems, multi-agent creative workflows, and cloud and data infrastructure. Each initiative has its own product requirements and success measures.

My ownershipProduct architecture, workflow definition, evaluation, platform contracts, and infrastructure priorities.

Read the case study: Making shared decisions useful. Making AI workflows reviewable.Staff Product Manager
Three disciplines

Shared decisioning, creative workflows, infrastructure

Separate initiatives with distinct product requirements
  • AI workflows
  • Shared decisioning
  • Infrastructure

How I think

The faster system is not
always the better product.

In Programmatic Loyalty, I initially wanted more responsive campaign decisions. Engineering helped expose the cost: less predictable spend behavior and a harder system to diagnose. I narrowed the scope to protect the advertiser promise.

How I led the work
One decision, in practiceProgrammatic Loyalty

How much should a live campaign react to new signals?

The promise
Let an advertiser set a spend goal and target return, then make campaign decisions they could trust.
The tradeoff
Reacting to fresh intent could recover delivery faster. It also made spend changes harder to explain and debug.
What I changed
Keep heavier optimization offline and the live path predictable. Improve the guidance before adding more live complexity.
Read the full decision

Built out of curiosity

Everyday friction.
Something worth exploring.

Outside work, I build around the small frustrations I notice in products I use. A working prototype helps me explore the interaction and make the tradeoffs concrete.

8 independent prototypes

WhatsApp Regional Voice

Language & access

Can playback keep its regional character?

Explore dialect detection, transliteration, and regional voice delivery through an AI pipeline.

Read the video description

A message from Rahul appears over a navigation-style interface: “Bhai, kaha pohcha?” The message asks where the recipient has reached. The preview demonstrates spoken playback from the notification. This is a concept recording, not a deployed driving interface.

AI integration prototype

The product thinking behind WhatsApp Regional VoiceExplore the interface: WhatsApp Regional Voice (opens in a new tab)
There is more on the workbench5 more explorations

Independent concepts, with sample data. Each story explains what works, what is simulated, and what I would evaluate next.

Experience

A career built
across the whole problem.

From analyzing advertising performance to owning products and leading shared platforms. Open a chapter for the roles, responsibilities, and outcomes.

Rakuten

May 2021 to present · San Mateo, CA

Advertising, recommendations, AI, and infrastructure

Staff Product Manager

June 2026 to present

Product architecture for multi-agent GenAI workflows, shared recommendation and ranking systems, and cloud infrastructure. Partnering with engineering and data science on the contracts and metrics that make platforms reusable.

Current scope; initiative timelines span my work at Rakuten

Senior Product Manager

July 2022 to June 2026

Launched Programmatic Loyalty and defined personalization, cohort targeting, and online/offline decisioning across Kafka, Redis, and Snowflake.

Programmatic Loyalty reached $500M+ in annual product revenue

Product Manager

May 2021 to July 2022

Launched Personalized Rewards, designed randomized experiments with CUPED, and partnered with data science on propensity modeling for merchants with sparse historical data.

Personalized Rewards and sparse-data merchant modeling were separate initiatives

LeanData

August 2020 to May 2021 · Santa Clara, CA

Workflow automation and revenue operations

Product Manager

August 2020 to May 2021

Shipped no-code rules and workflow orchestration. Built real-time lead routing and led the integration of predictive scoring into production CRM workflows.

40% fewer configuration escalations · 21% higher conversion through predictive scoring

Media.net

August 2017 to July 2019 · Mumbai, India

Advertising economics and performance analysis

Product Analyst

August 2017 to July 2019

Built Python regression models to connect campaign performance with traffic economics and translate analysis into bidding and traffic decisions.

Contributed to a 20% increase in ad revenue

Education

Master of Science in Statistics
University of Maryland, College Park
2019 to 2020
Bachelor of Engineering in Computer Science
University of Mumbai
2013 to 2017

Beyond the roadmap

Curiosity has
a few other outlets.

My reading shelf moves between science fiction, science, history, and mysteries. The playground is another place I follow an idea far enough to make it playable.

Browse my reading shelfVisit the playground
The Hitchhiker’s Guide to the GalaxyThe BodyThe Wager
A few books from my shelf

Get in touch

A conversation is a good
place to start.

I am interested in product leadership roles that bring customer understanding and technical depth together. If that is what your team needs, I would enjoy comparing notes.

Based in Sunnyvale, CA.
Open to remote, hybrid, and relocation.

deopranav2808.work@gmail.com
LinkedInRésumé (PDF)

Your message is delivered through Formspree. Prefer email? Use the address alongside this form.