Brain DumpProjects

Pooja Singh

Full Stack Engineer

About Me

Hi there, I'm Pooja – a caffeine-fueled explorer of all things tech! I moved from India to the US in 2016 to pursue my MS at USC, and since then, I've been on a mission to automate all things boring—because who wants to waste time on mundane tasks? With over 7 years of experience as a Full Stack Engineer, I've dabbled in everything from frontend wizardry to backend sorcery, cloud technologies, and even applied ML.

Lately, I've been diving deep into the world of Large Language Models (LLMs)—it's an exciting time to be a developer! I spend my free time building applications that leverage the power of LLMs, transforming complex ideas into simple, user-friendly solutions. 🚀

When I'm not coding, you'll find me outdoors— whether it's hiking, backpacking, scuba diving, bouldering, or planning my next road trip. I have a goofy Golden Retriever, Ralph, whom I love and adore.

I enjoy connecting with new people and building relationships. So if you're looking for someone who's passionate about building innovative applications and enjoys a good cup of coffee (or three), let's connect!

Tech Stack

ReactNode.jsTypeScriptNext.jsAWSPythonLLMsMLAngularSQL/NoSQLJestReact Testing LibraryJUnitPlaywrightPuppeteerAzureLLM ApplicationsPrompt EngineeringRAGVercelModel Context ProtocolOpenAI RealtimeVoice AIGenerative AIPostgreSQLVectorDB

Recent Projects with LLMs

Audentic – Voice AI Integration Platform

Built a plug-and-play platform for embedding real-time voice assistants into websites, abstracting speech recognition, NLP, and TTS into a unified end-to-end model. Enabled seamless integration of OpenAI's Realtime API for multimodal AI interaction with minimal developer overhead.

OpenAI RealtimeVoice AIModel Context ProtocolNext.jsTypeScript
Detextit.Music: AI-Powered MIDI Generation

Created a full-stack music generation application using Claude Sonnet-3.5 and Haiku-3 LLMs for multi-track MIDI compositions. Implemented function-calling and advanced prompt engineering.

ClaudeMIDITypeScriptFunction CallingPrompt Engineering
Detextit.3D - AI-Powered Blender 3D Scene Generation

Developed a Blender plugin for AI-assisted 3D scene generation using Claude LLM. Implemented automated asset selection and arrangement by indexing user's assets and free asset libraries.

BlenderClaudePython3D ModelingAsset Management
Therapeutic Conversational AI

Engineered prompts for Claude-powered interactive therapy, focusing on user experience and therapeutic appropriateness. Implemented evaluations and guardrails to ensure the chatbot remains within the therapy domain.

ClaudeTherapyGuardrailsGPT-4Evaluation
NLP Learning Assistant with Adaptive Instruction

Built a RAG-based educational assistant using Hugging Face NLP course materials. Created an adaptive learning system that personalizes instruction based on user interaction patterns.

RAGHugging FaceNLPEducationAdaptive Learning
Career Journey
  1. Software Development Engineer 2, Microsoft

    Redmond | Feb 2020 - Present
    • Azure Analytics - Lens Explorer
      • Developed key features for Lens Explorer data analysis and visualization service, enabling users to connect to their data sources and extract meaningful insights
      • Contributed to Orchestrator UX for executing scripts, transforming data across databases, and designing job scheduling features for seamless data workflows
      • Built comprehensive test infrastructure platform including unit, integration, and end-to-end tests, promoting testing culture across the team
      • Led adoption of Azure Managed Grafana for dashboards, creating migration tools to convert existing dashboards to Grafana-compatible formats
      • Enhanced ElixirFX internal design framework through bug resolution and performance optimization, ensuring smooth customer transition
    • Azure Monitor – Data Collection Rules
      • Redesigned and migrated Azure Monitor's Data Collection extension from KnockoutJS to React, significantly improving usability and maintainability
      • Designed and enhanced Data Collection Rules to standardize workflows across Azure Monitor components, addressing customer pain points and driving significant improvements in customer satisfaction (CSAT)
      • Architected an AI-powered agent to eliminate accessibility regressions by integrating automated audits with prompt workflows and Model Context Protocol (MCP) servers, enabling autonomous detection, resolution, and deployment of fixes to production
      • Led a cross-team (v-team) initiative to scale test automation across 5+ product teams, leveraging large language models (LLMs), Playwright MCP server, and AI-assisted test generation techniques to auto-generate tests and significantly increase coverage for mission-critical codebases
  2. Sr. Software Engineer, Visa Inc

    Denver | March 2018 - Feb 2020
    • Infrastructure Architecture & Engineering
      • Supported Visa's Continuous Availability Program through solution design and IT disaster recovery initiatives, improving system resilience improving uptime to 99.99%
      • Developed a proprietary system to measure and automate critical infrastructure processes, leading to increased efficiency and faster delivery
      • Implemented the Central Infrastructure Service Request Portal and IT Disaster Recovery portal, streamlining service requests and ensuring comprehensive recovery plans
    • Tech Strategy, Governance & Disaster Recovery
      • Re-engineered Visa's Technology Governance program into an internal SaaS solution for enhanced policy enforcement
      • Automated governance workflows across Visa's platforms, significantly reducing operational overhead
      • Built and optimized key features within the Technology Governance Portal for improved usability
      • Collaborated with stakeholders to iteratively enhance portal functionality based on user feedback
  3. Software Engineer Intern, Visa Inc

    Denver | May 2017 - Aug 2017
    • Tech Strategy, Governance & Disaster Recovery
      • Designed and developed Data Integrity dashboards ensuring policy compliance for Visa's Disaster Recovery
      • Automated IT Disaster Recovery exercise workflows to boost efficiency across the organization
      • Built Service Impact Assessment (SIA) tool providing risk and resiliency insights for Visa services
  4. Software Engineer, Capgemini Technology Services

    Bangalore, India | June 2015 - June 2016
    • Enterprise Cloud Solutions
      • Automated CI/CD pipelines using AWS, Azure, Chef, and Puppet to enhance deployment efficiency
      • Designed client cloud architectures and facilitated migrations from on-premises to Public and Hybrid Cloud environments