Karthika Boobathy
Role
Lead Product Manager
Based in
San Francisco Bay Area
Experience
10+ years in product
Availability
Open to Lead PM roles in AI-powered shopping, search and discovery, in the SF Bay Area or remote.

I turn customer reviews and marketplace signals into AI that helps people buy with confidence, in shopping and travel.

I'm a Lead Product Manager at TikTok, where I own the global shopping evaluation roadmap and work hands-on with ML engineers on generative AI for reviews and recommendations. Before that I spent three and a half years on hotel, flight and cruise search at Expedia Group, after building supply-chain products at a startup.

See selected work

Outcomes

Where
TikTok
When
2023–present
Role
Lead PM, owned roadmap
Basis
Cumulative, relative change. Absolute values confidential.

Gross merchandise value, TikTok Shop

Cumulative: +2.3% from generative AI review ranking, +1.5% from review insights, both on multi-modal LLMs. I defined the prompts, success metrics and evaluation loop with ML engineers, and diagnosed ranking failures by hand.

Where
Expedia Group
When
2020 → 2021
Role
PM, owned roadmap
Basis
Indexed to the prior year. Absolute values confidential.

Customer retention, opaque hotels

I ran cohort analysis to find where repeat engagement broke down, then set the acquisition and activation roadmap around what it showed.

Where
Elementum
When
2017–20
Role
PM, founding team
Basis
For a single client; not a portfolio-wide figure.

Expedite and handling costs

Automated anomaly detection and fallback orders, triggered by supplier misses and disruptions, for a single enterprise client. Built from scratch as a founding member of Elementum's Product Availability suite. Unplanned stockouts and manual escalations fell 23%.

Selected work

  1. Generative AI

    TikTok Shop

    2023–present

    Turning low-effort shopper reviews into a signal people trust enough to buy on

    Multi-modal LLM review ranking, summaries and insights, built with ML engineers on an evaluation pipeline grounded in hand-diagnosed ranking failures.

    +3.83% GMV, −3.54 pts size-related returns

  2. Retention

    Expedia Group

    2020–22

    Getting opaque-hotel customers to come back when repeat engagement was low

    Cohort analysis to find where repeat engagement broke down, then an acquisition and activation roadmap built on it.

    2× customer retention, year over year

    A hotel comparison screen on Hotwire matching a shopper's selected hotel to a cheaper, unnamed 'Hot Rate' room of the same class and location.
    The deal-match screen: the same 4-star Union Square hotel, priced lower when its name stays hidden until booking.

  3. 0→1, enterprise

    Elementum

    2017–20

    Catching supplier misses before they became stockouts for enterprise CPG brands

    Automated anomaly detection and fallback orders, built from scratch as a founding member of the Product Availability suite.

    $8.1M saved for one client, 23% fewer stockouts

    A network inventory chart showing available stock declining over several weeks into a projected negative balance, with first-outage and total-demand-at-risk figures above it.
    The network view: how far out a shortage will hit, and how much demand is at risk, before it becomes a stockout.
    An inventory detail page for a single product at one location, showing available and unavailable stock, a 4-week shortage forecast, and incoming and outgoing orders.
    The item-level detail a fallback order gets triggered from.

Also shipped

  1. TikTok
    2023–present

    Agentic review follow-up. An LLM-driven workflow that turns low-effort reviews into detailed, decision-useful ones. Average review depth rose 36.4% while review incentive spend fell 12%.

  2. TikTok
    2023–present

    AI review-writing guidance. Guidance that helps shoppers write reviews other shoppers can use, with better-targeted incentives. Useful-review coverage on key products rose 0.7% and GMV rose 1.1%.

  3. TikTok
    2023–present

    Cart-bundling recommendations. Product recommendations built on retrieval, embedding analysis and behavioural signals. Average order value grew 1.3×.

  4. TikTok
    2023–present

    European product-page launches. Launched compliance and safety details on the product page across 16 European countries (excluding the UK), scoping constraints with legal, safety, suppliers and marketing. User engagement rose 0.7%.

  5. Expedia
    2022–23

    Cruise search and evaluate revamp. Rebuilt cruise search around structured content. Churn fell 5.1% and conversion rose 4.3% year over year.

  6. Expedia
    2022–23

    Flight search ranking. ML ranking models for flight search, where I worked with data science on the objective function and constraint trade-offs. Conversion rose 3.21%.

    A recommended departing flights list, each option showing price, duration and stops.
    The recommended-flights view the ranking model feeds.

  7. Expedia
    2022–23

    Review-based hotel classifier. A multilabel classifier trained on customer reviews to categorise supply for search and comparison. Click-through rose 5.7% and conversion 1.09%.

  8. Expedia
    2022–23

    Recommended hotels. Hotel recommendations rebuilt for better match quality. Revenue per user grew 1.5×.

  9. Expedia
    2020–21

    Map-based hotel search. Search UX built around an interactive map, for better supply and demand matching. Conversion rose 2.7%.

    A hotel search map with two neighbourhoods selected and a results count at the bottom.
    Selecting a neighbourhood to search by area rather than by hotel name.

  10. Expedia
    2020–21

    Opaque hotels on Trivago. A new acquisition placement for opaque hotels on Trivago metasearch. Transactions grew 11% year over year.

  11. Elementum
    2017–20

    Cross-industry feature set. Use-case discovery across industries, a pitch to the C-suite, and adoption across the customer base. Daily active users rose 9%, licensing revenue 31% and NPS 17 points.

Contact

Availability
Open to Lead PM roles in AI-powered shopping, search and discovery, in the SF Bay Area or remote.
Work rights
US permanent resident. No visa sponsorship needed.

Sent to my inbox through Netlify Forms.

LinkedIn