Siddhant Raje: Building agentic AI systems and backend workflow platforms

M.S. Software Engineering graduate from San Jose State University. I build systems that ingest messy enterprise data, retrieve the right context, reason over operational risk, involve humans when needed, update systems safely, and leave clear audit traces.

SJSU M.S. Software Engineering
Agentic AI LangGraph, tool calling, RAG
Backend FastAPI, Kafka, PostgreSQL, AWS

Email Me Download Resume View Projects

99.9% uptime maintained on production services
100k+ monthly users supported by backend systems
1k+ concurrent sessions handled in real time
60% faster document triage workflow

Engineering Profile

Agentic AI / Backend Systems

Governed AI workflows built like production software.

I connect agents, retrieval, APIs, approvals, audit trails, and operational dashboards into systems teams can inspect and trust.

01 Agentic Workflows LangGraph · tool calling · approvals · evals
02 Context & Retrieval RAG · pgvector · FAISS · BM25
03 Backend Platforms FastAPI · Kafka · PostgreSQL · AWS
04 Production Control OpenTelemetry · retries · CI/CD · Docker

FOCUS Agentic AI and backend engineeringAgentic systems, backend platforms, and workflow automation.
STACK Python, FastAPI, LangGraph, ReactPlus MCP, RAG, Kafka, PostgreSQL, Docker, AWS, and observability.

Experience across AI workflows and production systems.

Work across agentic automation, document-processing infrastructure, distributed backend services, API performance, observability, and applied AI systems.

Nov 2025 - May 2026

Research Assistant

San Jose State University

Built a FastAPI document-processing backend with asynchronous conversion workers, PostgreSQL/pgvector metadata, and MinIO object storage to convert unstructured enterprise documents into AI-readable Markdown.
Designed a reliability-focused PDF extraction pipeline using typed OCR controls, output validation, and repair loops, improving Qwen OCR benchmark performance from 52.3% to 72.8% on olmOCR-Bench.
Developed tool-calling AI agents over processed PDFs and files using hybrid retrieval, citation grounding, and retrieval evaluation to answer document-specific questions with traceable source context.

Aug 2024 - May 2026

M.S. Software Engineering Graduate

San Jose State University

Focused on software architecture, distributed systems, cloud engineering, agentic AI systems, and full-stack product development.
Built project work around governed digital workers, developer platforms, reliable AI control planes, retrieval systems, and applied ML systems.

Jun 2022 - Jul 2024

Software Engineer

Tata Consultancy Services Ltd.

Owned backend modules for a Real-Time Income service, implementing Kafka producers/consumers and REST endpoints for income and household-status updates.
Optimized 20+ REST APIs with caching and fewer redundant downstream calls, lowering p95 latency from roughly 750 ms to below 500 ms while supporting 200-500 RPS production traffic.
Added OpenTelemetry tracing, correlation IDs, and CI/CD tests across distributed service-to-service and Kafka workflows, reducing MTTR by 10% and regression risk during releases.

Jul 2020 - Jul 2021

Machine Learning Engineer Intern

Resolute AI Software Pvt. Ltd.

Developed Python audio-processing pipelines using LSTM and Transformer models to analyze sentiment in 1,000+ customer voice interactions per month.
Integrated Kafka-based asynchronous processing into a TensorFlow object-detection pipeline, improving processing efficiency by 25% and reducing blocking work in inference.

May 2022

B.E. Computer Engineering

Don Bosco Institute of Technology

Built a foundation in software engineering, algorithms, databases, systems, and applied machine learning.

What I optimize for

I build AI systems like production systems: bounded, observable, auditable, and designed for real operational handoffs.

Governed automation Human-in-the-loop gates, policy-aware execution, retries, idempotent writes, and clear ownership boundaries.

Agents with evidence Retrieval, tool-use planning, scenario-based evaluations, execution traces, and citation-grounded context.

Production backend depth FastAPI services, Kafka workflows, PostgreSQL data models, OpenTelemetry tracing, CI/CD, and cloud deployment.

Agentic AI projects with platform depth.

Each project is framed around workflow ownership, retrieval, control points, operational visibility, and shipped behavior.

Agentic Operations

ProcureOps AI

Governed supplier-order digital worker for X12 confirmations, PO matching, risk investigation, approval routing, ERP-style updates, and audit traces.

FastAPILangGraphReactPostgreSQL

Repo

AI Control Plane

Agent Reliability Platform

Multi-tenant backend platform for versioned workflow execution, tenant-scoped access control, audit logging, evals, and rollouts.

PythonFastAPIPydanticPostgres

Repo

Retrieval App

Real-time Online Buying Assistant

Product research RAG application that ingests product pages and reviews to answer shopping questions with evidence-backed citations.

FastAPIStreamlitFAISSBM25

Repo

Document Platform

AI-Readable Document Processing

Dockerized FastAPI platform that converts PDFs, images, Office files, and text into Markdown for downstream AI retrieval.

FastAPIpgvectorMinIOOCR

Repo

Developer Platform

GitOps Developer Platform

Internal developer platform with Backstage scaffolding, GitHub Actions CI, Argo CD delivery, and Kyverno guardrails.

BackstageArgo CDKubernetesKyverno

Repo

Computer Vision

Toll Management System

Computer vision pipeline for real-time license plate detection, tracking, OCR, event queues, and billing automation.

YOLOv4DeepSortAWS SQSS3

Repo

Realtime Systems

Scalable Chat Platform

Responsive messaging platform with authentication, typed socket events, low latency delivery, and horizontal scaling.

Node.jsSocket.IOMongoDBDocker

Repo

Agentic AI stack, backend depth.

Grouped around agentic workflows, retrieval, backend services, distributed systems, cloud delivery, and production engineering.

Backend and Workflow APIs

Designing services and orchestration boundaries that are understandable, testable, and operationally reliable.

PythonFastAPIPydanticJavaC++JavaScriptTypeScriptNode.jsDjangoFlaskRESTpytestSQL

Agentic AI and Retrieval

Building bounded agent workflows with grounded context, tool calling, indexing, reranking, approvals, and evaluation loops.

RAGLangChainLangGraphMCPTool CallingFAISSBM25Information RetrievalNLPLLMs

Cloud and DevOps

Shipping through repeatable paths with CI/CD, containers, observability, policy, and infrastructure visibility.

AWSGCPDockerKubernetesKafkaJenkinsCI/CDDistributed Systems

Databases and Storage

Modeling data for workflow state, transactional backends, retrieval systems, caches, and object-backed document processing.

PostgreSQLpgvectorSQLiteCassandraMongoDBRedisMinIO

I am looking for teams building agentic AI systems, workflow automation, and reliable backend platforms.

Best-fit roles: agentic AI engineer, backend engineer, AI platform engineer, workflow automation engineer, full-stack platform engineer, or product-focused infrastructure engineer. I am available for full-time software engineering opportunities.

Email Me LinkedIn GitHub