Engineering AI
people can trust.

About

Full-stack engineer focused on RAG, agentic workflows and the backend systems that keep them accurate. Shipping software since 2021.

Currently

Software Engineer, AI at Agadia Systems, leading RAG chatbots and AI agents for the engineering lifecycle.

Enterprise RAG at Agadia Systems, 2026: by the numbers

technical documents indexed
170K+
faster knowledge lookup
60%
fewer hallucinations
50%
citation-backed answers
100%

Experience

Four teams · enterprise, startup & SaaS

2026–Now

Agadia Systems

Software Engineer, AI · New Jersey

Building the AI layer of Agadia's internal knowledge and SDLC platform: a RAG assistant over 170,000+ documents, agents that turn Jira change requests into requirements and test suites, and a pipeline that documents codebases on its own.

5–10 minper FRD, down from hours
2025–2026

Subscrbe AI

Founding AI Engineer · San Francisco

First engineer shaping architecture for a Flutter app on iOS and Android. Moved deep linking to native Universal and App Links and led RAG and agentic AI features.

+50%deep-link reliability
2024–2025

WashMetrix

Software Engineer · San Francisco

Built real-time React analytics for IoT laundry operations, Python and Node.js microservices, and automated ETL from POS systems into PostgreSQL. Took test coverage to 90%.

+25%user engagement
2021–2023

Zemoso Technologies

Software Engineer I · Hyderabad

Shipped React and GraphQL features, CI/CD with Docker, and auto-scaling workloads on AWS EKS. Mentored three junior developers.

−40%development cycle time

AI work at Agadia

Production systems, used by engineering, BA and QA teams

Case A · Internal tooling

Code-to-Docs: documentation that writes itself

Give it a Git repository and it returns Architecture, Technical and Functional documents ready for the wiki. Static parsers turn the code into AST JSON, the LLM writes each document in a single pass, and a knowledge graph we built of the system validates what it says.

4 parsers
Angular, C#, Python and SQL, traced from UI to database
1 pass
Each document generated straight from the AST JSON
Graph-checked
Output validated against a knowledge graph we built
3 docs
Architecture (C4), Technical and Functional, per repository

Claude, tree-sitter, Roslyn, Python ast, knowledge graph, Pydantic

Case B · Internal SDLC platform

From Jira ticket to tested feature, with agents

AI workflows that turn a Jira change request into a Functional Requirements Document, then a test plan, a validated test-case suite and an automated run, with people approving what the model proposes.

5–10 min
To draft an FRD that used to take hours
≥ 80%
Requirement coverage enforced by a validator agent
Human-approved
Agent edits are proposals; content hashes block conflicts
Streaming
SSE output and background jobs with live progress

Claude, Azure OpenAI (GPT-4o), LangChain, MCP, Playwright, FastAPI, Angular, PostgreSQL

Research

Applied software research · CSU East Bay

2023–2024

VR patient-safety trainer

Original developer · Lead author

Built a Unity app for Meta Quest 2 that puts nursing students inside 360° captures of CSU East Bay's simulation lab, shot on an Insta360 X2. Students search the room for 10 hidden safety hazards, with scoring, pop-up explanations and audio feedback for each find.

10
hidden hazards to find
1st
author, Springer CCIS
CSR
Scholar, CSU East Bay

Advancing Nursing Education Through Virtual Reality Training

CSCE 2024, Las Vegas · Springer CCIS vol. 2260, 2025

Read paper Watch the talk Unity, C#, Meta Quest 2, 360° capture
2025–2026

Standalone VR pilot with nursing students

Co-author · Handed over and mentored the codebase

Handed my codebase to the next student team and helped them get it running standalone on the headset, now with 5 hazards, a 5-minute timer and an end-of-session summary. Rejoined to co-author the pilot study and wrote its conclusions on challenges and next steps.

14
nursing students in the pilot
100%
found at least 3 of 5 hazards
4.64/5
want to use VR training more often

Improving and Piloting the Application of 360° Virtual Reality Technology in Nursing Education

CATA 2026, Honolulu · Springer CCIS, 2026

Read paper Unity, Meta Quest 2 standalone, User study
2026

Peer reviewer, JMIR Formative Research. Reviewing early-stage digital-health software studies.

Selected projects

Side projects, built end to end

No. 01 · Open source

DevTeam AI

A local-first team of AI agents that takes a feature request and a repository, then plans, designs, patches, tests and reviews the change, looping until every quality gate passes. Agents edit through patches, never a raw shell, and checks run in a locked-down Docker sandbox.

4 of 4 demo tasks shipped green, each in two iterations or fewer.

LangGraph, Ollama, FastAPI, Pydantic, Docker, pytest, Ruff, mypy, Bandit, Streamlit

No. 02 · Personal project

LaTeX Résumé Tailor

Works directly on your .tex source: pick a model, set tone and length, and get a keyword-optimized résumé with a live PDF preview in about two minutes.

Vector DB, RAG, OpenAI / LLMs, React, Node.js, LaTeX, PDF.js, Docker

Toolkit

AI / LLM
RAG, Agentic AI, Azure AI, pgvector, Pinecone, Gemini, OpenAI
Backend
Python, FastAPI, Node.js, Express, Java, Spring Boot, GraphQL
Frontend
React, TypeScript, Next.js, Angular, Flutter, Tailwind CSS
Data
PostgreSQL, MongoDB, MySQL, Redis, DynamoDB, Firestore
Cloud
AWS, Azure, Docker, Kubernetes, CI/CD, Serverless
Quality
Jest, Pytest, Postman, Git, Jira, Figma

Education

M.S. Computer Science

California State University, East Bay

2023 – 2025 · GPA 3.9 / 4.0

B.Tech Computer Science

CMR Institute of Technology

2018 – 2022 · GPA 8.8 / 10

Let's build something useful.

sailokeshreddy@gmail.com

Open to software engineering and AI-focused product work: roles, collaborations or a quick chat.

Or send a note