Vijaya Varala— AI PM
About

Engineer-turned-PM. Now AI-native.

I build enterprise data and platform products for high-stakes regulated markets: energy intelligence, financial data, smart mobility, and connected vehicles, always at the API and platform layer. Since 2024 I've paired that with hands-on AI execution. LLMs, RAG, agentic workflows, and voice agents shipped in production.

My story

I started as a software engineer, shipping systems at GM OnStar that pulled live NOAA weather feeds into ESRI mapping so call-center reps could find a stranded driver faster. That early lesson, that real users behave nothing like the persona deck, has driven every product decision since.

Over 8 years as a PM, I've owned developer platforms for 4,000+ engineers at S&P Global, modernized a chemicals analyst tool at IHS Markit, and led payments and platform strategy at Verra Mobility, including the Stripe integration that anchored Hertz's $177M rental-payments program.

In 2024 I made a deliberate bet on AI. As founding PM at AI Growth Vector, I've shipped agentic AI co-pilots, RAG service tools with hybrid retrieval and LLM-as-judge observability, and voice-agent provisioning systems for enterprise clients. The bet paid off. I now bring both deep regulated-market product leadership and hands-on AI execution to the table.

I optimize for impact-per-quarter, not seat-warming. At Verra Mobility I closed Hertz × Stripe and a $12M cost-savings program in under two years; at AI Growth Vector I've shipped 5+ enterprise AI products in the same window.

About the recent moves

Two questions sometimes come up about my recent path. I want to address them head-on because they're the wrong read.

AI Growth Vector is a real, full-time role. It's a productized AI consultancy where I'm the founding PM. I run the PRDs, the AI architecture trade-offs, the client roadmaps, and the post-launch metrics. The outputs are shippable enterprise systems: the DGA Systems call-center automation platform (~7,800 voice agents/year capacity) and the Automotive Dealership AI Service Co-Pilot (RAG + agentic). Not deck-ware.

Career arc

  1. 2024 — Present
    Founding AI, Product Manager · AI Growth Vector

    Productized AI consultancy. Led product on an end-to-end AI Service Co-Pilot for connected vehicles. CAN-bus telematics (brake wear, voltage, squeal frequency) stream into an Azure Databricks medallion lake. An anomaly-detection model scores them, with drift monitoring watching for shifts. When something crosses the line, an LLM agent (GPT-4o) takes over: it checks parts inventory, schedules a technician, and opens the CRM ticket. A RAG layer over the service manuals runs hybrid search (semantic + BM25) with parent-document retrieval, and an LLM-as-judge harness cut hallucinations 25%. My job was the calls only product could make: high recall on brake safety because a missed failure is unforgivable, high precision on oil-change nudges because nobody wants to be pestered, and a feedback loop that keeps the model honest over time.

  2. 2022 — 2024
    Product Manager · Verra Mobility

    Owned platform modernization and payments. Secured a $177M Hertz rental-payments contract via Stripe integration; stood up a production-like staging environment; led API & tech-debt strategy across 70+ platforms.

  3. 2021 — 2022
    Principal Product Manager · S&P Global Market Intelligence

    Owned the 12-month strategic narrative for a global developer platform serving 4,000+ users. Led the AWS migration (−20% opex), standardized the post-merger DevOps stack across 15 services and 50K+ users, and ran the first OCTO Dev Tools survey. The workflow fixes that came out of it cut incident escalations 60%.

  4. 2017 — 2021
    Technical Product Manager · S&P Global / IHS Markit

    Product for the chemicals data platform. Reduced time-to-insight by 50% by eliminating data fragmentation, led the legacy-to-modern migration with a 70% performance lift, and shipped 100% feature parity through a high-stakes platform transition aligned to a nine-figure business unit.

  5. 2013 — 2017
    Senior Software Engineer · S&P Global / IHS Markit

    Engineered high-scale ETL consolidating 3 legacy systems for the annual Chemical Supply & Demand balancing; modernized legacy codebases that lifted pricing-analyst efficiency 80%; stood up Elastic ELK observability across global data products.

  6. 2009 — 2012
    Applications Developer · iOffice Corp

    Re-architected a ColdFusion application to Java/Hibernate, lifting processing speed 45% and stabilizing a core enterprise workflow.

  7. 2007 — 2009
    Systems Engineer · GM OnStar

    Shipped Subscriber Crisis Assistance at GM OnStar. NOAA weather feeds plus ESRI ArcGIS mapping in Adobe Flex on BEA WebLogic, giving call-center reps real-time situational awareness for stranded drivers. First production exposure to connected-vehicle engineering, and the foundation under every PM decision since.

Education & training

  • MS, Computer Science
    University of Texas at Arlington
  • Product Strategy, Executive Education
    Kellogg School of Management

Want the full story over a call?

Get in touch →