LIVE NOW: Building Dynamic Ads at Meta · San Francisco · open to chat EST. 2014

Career · 2014 → now

Meta '24 → now Stellar Foundation '23 – '24 Robinhood '21 – '22 Dropbox '19 – '21 Quantstamp YC W18 '17 – '18 Barclays '15 – '17

Shipped · by the numbers

$7M+
annualized iRev
Meta · frequency capping
−48%
RPC latency
Stellar Platform
30+
product flows orchestrated
Robinhood · SMaaS

Featured · now

Frequency Capping Platform · Meta Ads

Policy-driven, low-latency capping for Dynamic Ads. Combines onsite engagement with offsite conversions via batch and near-real-time streaming. Integrated into the final-stage Ads ranking stack without impacting serving SLAs. $7M+ iRev delivered →

Also built

LLM Ads-debug agent

Internal AI agent using RAG over logs/metrics + tool use to cut root-cause time for Ads delivery issues.

State Machine as a Service

JSON-configured workflow engine at Robinhood. 30+ flows; new launches 3–5× faster.

Stellar RPC service

Re-engineered transactions/events handling with optimized caching. Core endpoint latency down 48%.

Dropbox Spaces

Led 4-engineer team building core collaboration features: notifications, membership, workspace creation.

Full experience

Oct 2024 → now

Meta

Senior Software Engineer · Dynamic Ads

  • Led architecture and development of a scalable frequency capping platform for Dynamic Ads within Meta's Ads recommendation stack, combining onsite engagement with offsite conversions via batch and near-real-time streaming pipelines.
  • Built a generic, policy-driven framework supporting per-user, per-ad, and per-campaign caps with dynamic tuning based on objective, cohorts, and signal quality.
  • Integrated capping into the final-stage Dynamic Ads recommendation stack as a low-latency ranking and delivery-control feature without impacting serving SLAs.
  • Delivered $7M+ annualized iRev by reducing inefficient repeat exposure and improving delivery quality.
  • Built an internal LLM-powered AI agent for Ads debugging — retrieval over logs/metrics, prompt orchestration, and tool use to accelerate root-cause analysis.
Jun 2023 → Oct 2024

Stellar Development Foundation

Senior Software Engineer

  • Architected and developed a new backend system for handling transactions and events in the Stellar RPC service.
  • Re-engineered the RPC service with optimized caching in Stellar Platform — core endpoint latency down 48%.
  • Led platform and SDK releases for enterprise deployments and global developer adoption across the Stellar ecosystem.
Nov 2021 → Oct 2022

Robinhood

Senior Software Engineer

  • Architected and implemented a State Machine as a Service platform (Python, Go, Redis, PostgreSQL, AWS) — engineers deploy complex product flows via JSON in ~10 minutes instead of multi-week custom builds.
  • Built the core execution engine, transition framework, caching layer, and durable state store for UI-driven workflows with backend DB/RPC side effects — cutting duplicated product-flow engineering by ~42%.
  • Drove adoption across 30+ product flows — crypto asset launches, onboarding, eligibility checks, multi-step transaction experiences — accelerating new product launches 3–5×.
Jun 2019 → Nov 2021

Dropbox

Software Engineer · Dropbox Spaces

  • Designed and built core collaboration features for Dropbox Spaces, a project-centric virtual workspace for teams (Go, Python, Protobuf, gRPC).
  • Led a 4-engineer team through design reviews, sprint planning, on-call, and cross-functional execution across notifications, membership management, and workspace creation.
  • Developed backend services and APIs integrating with Dropbox identity, sharing, and collaboration systems.
Jul 2017 → Aug 2018

Quantstamp · YC W18

Founding Software Engineer

  • First engineer on the web product team — built the MVP from scratch.
  • Worked directly with the CEO and CTO to build an asynchronous backend so customers could receive audit reports without blocking on submission.
Mar 2015 → Jul 2017

Barclays

Senior Software Engineer · Payments / Fraud / Collections

  • Developed modules in Payment, Fraud, and Collection services handling millions of user requests daily (Java, Spring, AWS).
  • Optimized the transaction/payment validation pipeline — 30% latency reduction via async processing and batch handling.
  • Collaborated with the data science team on fraud detection models — +25% detection accuracy.
  • Implemented rule-based decision engines to prioritize collection efforts by customer profile and payment history.

What people I've worked with say

"Prit was great to work with. He thought thoroughly through his eng designs so we were prepared for complications up front and brought wisdom to our solutions. He also dove deep into difficult problems and worked independently, asking good questions when appropriate. Prit was also a positive influence on the team as he built meaningful personal connections and always maintained an upbeat attitude even when times were tough or uncertain without being fake or disingenuous. This combination of technical expertise and positive influence makes him a great team member!"
— Co-founder & CTO @odo | Ex-Robinhood
"I was the mentor of Prit Sheth in Global Think Tank Innovation Contest held at Barclaycard and run by MasterCard. His ingenious idea of a Smart Recommendation Engine not only became the only project from India to qualify but also won the contest. He has proven himself as an exceptionally committed collaborator who consistently asks questions and seeks guidance when needed, to improve his understanding of the business, technology, and solutions. In the final round, he had pitched his idea to the CEO of Barclaycard US and other executives and Innovation Council. The presentation was very well received. In my time with Prit I observed him as forward thinking yet grounded in the technology that would both limit and enable the solution. He would be an asset to the program and I offer my highest recommendation."
— SVP Card Acquisition and Digital Product