Prachi Saibewar · Software Engineer

Building production backend, full-stack & applied AI systems.

Python/FastAPI · C#/.NET · React/TypeScript · Distributed Systems · RAG/Agents

View Projects Resume LinkedIn GitHub
3+ yrs
software engineering
300+
platform integrations
29%
lower p95 API latency
79% → 92%
grounded LLM answers

Currently: building IncidentIQ's approval API and React/TypeScript frontend · open to Software Engineer roles

Selected Work

IncidentIQ architecture: LangGraph agent loop, approval gate, PostgreSQL with pgvector

Featured · Agentic AI · Backend

IncidentIQ

An incident-response agent that helps on-call engineers diagnose production issues with evidence, and never changes production without human approval.

  • LangGraph agent with persisted state that survives process restarts
  • Hybrid retrieval: pgvector similarity + Postgres full-text, fused with reciprocal rank fusion
  • Write actions gated structurally, not by prompt; evaluation harness publishes every result
0.80 Recall@5 0.92 MRR
  • Python
  • FastAPI
  • LangGraph
  • PostgreSQL
  • pgvector
  • Docker

Approval API, React frontend, and tracing in progress.

Read case studyGitHub

Experience

Full resume (PDF) ↗
  1. Sep 2025 - Aug 2026 Rokt (mParticle)

    Software Engineer

    • ~30% faster identity investigations: Customer 360 / IDSync workflows in C#/.NET and React/TypeScript
    • 29% lower p95 latency: 185 → 132 ms using Redis caching and async parallel retrieval
    • 300+ integrations: consent-aware audience activation across a high-volume customer-data platform
  2. Jun 2025 - Sep 2025 University of Central Florida

    Research Engineer

    • 79% → 92% grounded responses: Python/FastAPI RAG on PostgreSQL/pgvector, tuned against a 300+ query benchmark
    • 98%+ schema-valid outputs: structured outputs with validation, retries, guardrails, and human review
    • Chartalist.org: rebuilt from legacy PHP to React/TypeScript
  3. Jun 2021 - Jul 2023 Infosys

    Software Engineer

    • 5,000+ daily transactions: C#/.NET REST services and SQL-backed onboarding and transaction workflows
    • Reconciliation incidents ~25 → 6 per week: Kafka sync across 5 systems with idempotent consumers and DLQs
    • 25% faster APIs: query, index, and data-access optimization

Open Source & Research

Writing

About

I'm a Software Engineer with 3+ years building production APIs, distributed integrations, full-stack products, and applied AI systems.

My background spans high-volume customer-data infrastructure at Rokt (mParticle), RAG and research tooling at UCF, and enterprise backend systems at Infosys. I'm strongest where backend engineering meets product delivery and applied AI. Based in Orlando, FL.

Prachi Saibewar

Skills

Backend & Distributed Systems
Python, C#, SQL, FastAPI, .NET/ASP.NET Core, Node.js, REST APIs, Microservices, Event-Driven Architecture, Kafka, Redis, Async Processing
Frontend & Product
React, TypeScript, JavaScript, Next.js, Redux, HTML/CSS
Data, Cloud & Reliability
PostgreSQL, MongoDB, AWS, Docker, Kubernetes, CI/CD, GitHub Actions, OpenTelemetry, Prometheus/PromQL, Chronosphere, Jest, Playwright
Applied AI
LLMs, RAG, AI Agents, LangGraph, LangChain, pgvector, Vector Search, LLM & Retrieval Evaluation, Structured Outputs, Tool Calling, Guardrails, Human-in-the-Loop

Education

Aug 2023 - May 2025 University of Central Florida

M.S. Computer Science, GPA 3.9

Aug 2016 - May 2020 Savitribai Phule Pune University

B.E. Computer Science

Get in Touch

Open to Software Engineer roles across backend, full-stack, and applied AI. Email me directly or use the form.

prachi.saibewar.cs@gmail.com