Hi, I'm Pooja 👋

Pooja Singhal

I build operating models for high-stakes AI/ML & platform programs.

Senior Technical Program Manager with 11+ years driving cross-organizational technology programs at Twilio, Affirm, and Amazon — with hands-on roots as a software engineer who shipped production Java systems. I turn ambiguous, multi-team initiatives into roadmaps, governance, and shipped outcomes in regulated, real-time, high-scale environments.

Pooja Singhal
$2B+ GMV enabled at peak scale
11+ Years leading tech programs
88% ML model accuracy delivered
77% Fewer false interruptions (VAD)

01. About me

I'm a Senior Technical Program Manager who thrives where programs are ambiguous, cross-organizational, and high-stakes. My specialty is building the operating model — the roadmap, governance, cadences, and decision frameworks — that lets many teams move as one.

Today I lead Conversational AI programs at Twilio, where I shape strategy for a real-time AI/ML platform and run ML model evaluations (precision, recall, false-positive rate) from POC through GA. Before that I scaled infrastructure for peak events at Affirm and led one of Amazon's most complex regulated launches, Amazon Pharmacy.

I started in engineering. At IBM and Adobe I designed, built, and shipped production Java/Spring applications and REST APIs — owning feature development, code reviews, and production support — and later certified production code as an Amazon Application Security Certifier. That hands-on foundation is why engineers trust me in the room, and why I can run ML evaluations and architecture sign-offs credibly. (This site is hand-coded, too.)

I work most often across:

  • AI/ML & GenAI platforms
  • Real-time distributed systems
  • API & data-platform migrations
  • Regulated environments (HIPAA, PCI, SOC 2, GDPR)

02. Where I've made an impact

Jan 2025 — Present Twilio

Senior Technical Program Manager — Conversational AI

Shaping strategy and the operating model for Twilio's real-time AI/ML platform (50K+ daily sessions), unifying 7 concurrent workstreams under one roadmap and aligning 5 engineering squads, data science, legal, and external vendors from roadmap to GA.

  • Drove 30% platform adoption growth across a unified roadmap.
  • Owned a $1.4M AI vendor transition for real-time spam detection — 88% accuracy (+28% vs. legacy), <5% false positives.
  • Led VAD model evaluation that cut false interruptions 77% and improved handle time ~22s/call.
  • LLM / GenAI
  • ASR · NLU · TTS · VAD
  • ML eval frameworks
  • Vendor POC
Nov 2023 — Dec 2024 Affirm

Senior Technical Program Manager

Directed peak-season infrastructure scaling for Amazon, Apple, Shopify, and Walmart partnerships — engineering systems to sustain 50x baseline traffic through Prime Day, iPhone launches, and BFCM.

  • Enabled ~$2B in GMV with zero critical incidents across 8 engineering teams.
  • Spearheaded a $12M cost-optimization initiative, delivering 6 weeks ahead of schedule.
  • Architected the org-wide OKR planning cadence with executive leadership.
  • Built a compliance governance framework that improved resolution efficiency 70%.
  • Capacity planning
  • Incident readiness
  • OKR / KPI frameworks
  • Compliance
Mar 2020 — Aug 2023 Amazon

Technical Program Manager II

TPM lead for the launch of Amazon Pharmacy — one of Amazon's most complex, regulated launches — orchestrating product, engineering, pharmacists, 3 vendors, and legal under strict HIPAA/PII/PHI requirements.

  • Delivered two fulfillment centers (Austin, Phoenix) 2.5 months ahead of schedule.
  • Architected a $10M/year platform migration (Snowflake→Redshift, Looker→QuickSight) across 20+ teams, ~800 tables, zero downstream disruption.
  • Used NLP root-cause analysis on JIRA data to cut vendor bug rate 30%.
  • Certified 3 internal apps as an Amazon Application Security Certifier.
  • HIPAA / PHI
  • Data-platform migration
  • Change-control governance
  • AppSec
Jul 2019 — Mar 2020 Vivint Smart Home

Project Manager / Scrum Master II

Led cross-functional delivery for the CoreApps vertical — driving root-cause analysis on critical system failures and introducing agile practices that scaled across the org.

  • Recovered a critical PIN/password failure to a 99% success rate via API retry logic, synchronous fallback, and idempotent request handling.
  • Introduced Scrumban, cutting delivery timelines 20% and defects 40% — then scaled it to 5 more teams.
  • API reliability
  • Root-cause analysis
  • Scrumban / Agile
Mar 2014 — Aug 2018 IBM

Senior Application Developer — Software Engineering

Designed, built, and shipped MobileBOND, IBM's enterprise self-service procurement platform — a Java (J2EE/Spring) web application with a Hibernate/JPA persistence layer, owning feature development, REST API services, code reviews, and production support.

  • Engineered the companion mobile app end-to-end — exposing catalog, approval workflows, and order tracking via Java/Spring REST APIs — cutting customer-care load 15% and delivering $96K in annual savings.
  • Built core backend services: role-based approval workflows, catalog/inventory modules, and order-fulfillment integrations; refactored legacy code and resolved production defects across the Java codebase.
  • Led a 6-engineer agile team through sprint planning and design reviews; automated regression testing to cut combined dev+QA cycle time 67%.
  • Java · J2EE · Spring
  • Hibernate / JPA
  • REST APIs
  • SQL
  • Test automation
Sep 2012 — Feb 2014 Adobe

IT Consultant — Software Engineering

Developed internal Java-based automation tools and backend services to streamline IT service operations, partnering with engineering teams on fixes, enhancements, and release coordination.

  • Built a Java ticket-classification engine that auto-routed requests by category — improving resolution speed 20% and driving $20K in annual savings.
  • Maintained Java/SQL backend scripts and reporting modules across the internal support application stack.
  • Java
  • SQL
  • Automation tooling

03. Selected highlights

88%

ML spam detection, re-platformed

Ran a 3-vendor POC on 10K annotated samples with go/no-go KPI thresholds, then phased rollout with bi-weekly feedback loops — +28% accuracy over a legacy rule-based system.

$2B

Peak-season scale, zero incidents

Capacity models, load-test cycles, and incident readiness across 8 teams sustained 50x baseline traffic through Prime Day, iPhone launches, and BFCM.

0→1

Conversation Intelligence platform

Led discovery, authored the PRD, and defined architecture for an AI analytics platform: generative insights, sentiment, entity recognition, and summaries.

2.5mo

Amazon Pharmacy, ahead of schedule

End-to-end technical enablement of two fulfillment centers under strict HIPAA/PHI, with operating procedures, change control, and SLA management.

70%

Compliance governance, automated

Designed an issue-management framework and audit dashboard with SLAs, escalation rules, and aging thresholds — real-time visibility into regulatory risk.

$12M

Infra cost optimization

Built the executive business case with quantified ROI across 3 orgs, negotiated reallocation, and accelerated delivery 6 weeks early.

04. How I work & what I know

Program Strategy & Governance

  • Cross-org program leadership
  • Influence without authority
  • Technical roadmap strategy
  • Operating cadence & governance design
  • Phased rollout & vendor migration
  • Risk / RAID management
  • OKR / KPI frameworks
  • Executive reporting & dashboards

AI/ML Program Delivery

  • ML model evaluation frameworks
  • Precision · recall · WER · FPR
  • AI vendor POC design & selection
  • Model deployment & phased rollout
  • LLM · GenAI · ASR · NLU · TTS · VAD
  • Responsible AI & compliance
  • Agentic AI program management

Technical & Data

  • Hands-on software engineering (Java / Spring)
  • Real-time distributed systems
  • API architectures & platform infra
  • Data-platform migrations & pipelines
  • SQL · Python · Java · J2EE / Spring · AWS
  • Snowflake · Redshift · Tableau · QuickSight
  • REST / WebSocket APIs · CI/CD
  • HIPAA · GDPR · SOC 2 · PCI

Certifications

  • SAFe 4.0 Agilist
  • PMI CAPM
  • Certified Scrum Master (CSM)
  • Certified Scrum Product Owner (CSPO)
  • Oracle Certified Java Professional 6
  • IBM DB2 9 Fundamentals
  • Amazon Application Security Certifier

05. Let's talk

I'm always glad to talk about AI/ML programs, platform scale, and how strong operating models ship hard things. Whether you're hiring, collaborating, or just curious — drop me a note.

or reach me directly