Skip to content
View poojas49's full-sized avatar

Block or report poojas49

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
poojas49/README.md

Pooja Shinde

Software Engineer with 4+ years of experience building high-throughput distributed systems. At Walmart Global Tech, shipped campaign management and reporting infrastructure processing 100M+ events/day with sub-15-minute SLAs and 200K+ external API requests/day. Currently finishing an MS in Computer Science at UIC (GPA 4.0) and applying those systems instincts to production AI — RAG pipelines, agentic workflows, and LLM inference.


What I Build

Backend & Distributed Systems

  • Event-driven pipelines, low-latency APIs, bulk processing, and observability
  • Java, Scala, Spring Boot, Akka, Kafka
  • Latency and reliability are constraints, not afterthoughts

AI Systems (Production, Not Demos)

  • Agentic RAG with LangGraph, vector search, guardrails, and human-in-the-loop escalation
  • LLM-based confidence scoring and authenticated multi-department access

Data Infrastructure

  • ETL pipelines, Spark/Hadoop at scale, MS SQL Server, PostgreSQL
  • Owned pipelines from ingestion through analytical delivery

Experience

Walmart Global Tech — Software Engineer II → III (Jul 2021 – Jun 2024)

  • 100M+ advertising events/day through async Azure Queue pipeline, <15 min SLA
  • 50M+ state changes/month in persistent campaign history system
  • Bulk APIs cutting partner API calls by 70%; 200K+ external requests/day under governance
  • 93% reduction in manual keyword reviews via Kafka event-driven automation
  • 40% reduction in MTTR via Prometheus/Grafana observability
  • Led PR standards for 8-engineer team; reduced escaped production defects by 30%

UIC Technology Solutions — Graduate Assistant (Oct 2024 – Present)

  • Agentic RAG chatbot serving 12+ university departments (LangGraph, FastAPI, vector search)
  • 87% improvement in response accuracy via query reformulation and optimized retrieval
  • Guardrails with LLM-based confidence scoring and automated escalation to human support

Projects

Distributed LLM Training & Inference System · Scala · Spark · DL4J · gRPC · AWS

  • Hadoop/Spark pipeline: ingestion → tokenization → embeddings → semantic similarity
  • gRPC inference layer routing across local Ollama and AWS Bedrock with multi-turn memory
  • Deployed as containerized microservices on EC2 with Lambda serverless routing

RhythPic — AI Music Visualization Platform · React · Node.js · MongoDB · Genius API

  • Full-stack app generating AI visuals synchronized to time-coded lyrics
  • MongoDB caching reduced external API calls by 60%
  • Real-time lyric sync and playback queue via React Context API

Stack

Java · Scala · Python · JavaScript · TypeScript · SQL
Spring Boot · Akka · Play · Node.js · Kafka · FastAPI · gRPC · REST
Hadoop · Spark · ETL · LangGraph · RAG · Vector DBs · LLMs
AWS · Azure · Docker · Kubernetes · GitHub Actions
Prometheus · Grafana · Splunk


Open to full-time roles in backend engineering, distributed systems, and AI engineering (Gen AI and agentic workflows).

Pinned Loading

  1. LLM-Data_Processing LLM-Data_Processing Public

    Scala

  2. LLM-Model-Training-Pipeline LLM-Model-Training-Pipeline Public

    Scala

  3. LLM-Conversation-Service LLM-Conversation-Service Public

    Scala

  4. Fashion_Inspo Fashion_Inspo Public

    Python