Skip to content
SK
Resume
Saketh KanchiFull-Stack AI Engineer · Jersey City, NJ

Portfolio // Full-Stack AI

Saketh
Kanchi

RAG systems, LLM integrations, and production full-stack features.

Email meResume ↗
GitHubLinkedIn

Scroll ↓

Selectedwork

01/ 05

drive-rag

Edge RAG System

BYOK Retrieval-Augmented Generation app that answers natural-language questions over the documents in a Google Drive folder. Ships edge-only on Cloudflare Workers, Vectorize, and Workers AI — no server to run for queries.

  • Cloudflare Workers
  • Workers AI (bge-base + llama-3-8b)
  • Vectorize
  • Google Drive API + OAuth
  • discord.js

GitHub ↗

Read the build →
02/ 05

Parley

Audio / Multi-process Bot

Open-source, self-hosted Discord bot that records voice channels, transcribes audio locally via a Python faster-whisper sidecar, and generates structured AI meeting notes. Privacy-first — no cloud audio upload.

  • Node.js + discord.js
  • Python + faster-whisper
  • FastAPI
  • node:sqlite + FTS5
  • Pluggable LLMs (Gemini/OpenAI/Ollama)

GitHub ↗Landing ↗

Read the build →
03/ 05

tracker

Developer Tool / CLI

One terminal command that shows the remaining quota on every AI subscription you own — Claude, Grok, Codex, Gemini, and OpenAI — instead of logging into five dashboards to find out who still has budget. Published on PyPI as ai-quota-tracker.

  • Python 3.11+
  • SQLite
  • OAuth 2.0 / OIDC refresh
  • httpx + rich
  • Discord webhook

5 providers, one dashboard

GitHub ↗PyPI ↗

Read the build →
04/ 05

Kitty — Visual Studio Code Theme

Developer Tool / Theme

A production color theme extension for Visual Studio Code, published on the Marketplace with a real user base and accessibility-driven iteration.

  • Node.js
  • VS Code Extension API
  • JSON token themes
  • vsce

0+downloads

Marketplace ↗

Read the build →
05/ 05

Multiple Disease Prediction

ML Comparison / Streamlit

A machine-learning web app that predicts the likelihood of multiple diseases — Diabetes, Heart Disease, and Parkinson's — from user-entered health metrics. Multi-page Streamlit interface over pre-trained ML models.

  • Python
  • Streamlit
  • Scikit-learn
  • Pandas + NumPy
  • Kaggle datasets

GitHub ↗

Read the build →

Experience

01
Fund Flow OSFull Stack AI EngineerDec 2025 – Present

Jersey City, NJ

  • Own customer-facing features end to end in React, Node.js, and REST APIs; integrated Apollo.io so the product auto-enriches and stores 500+ lead records.
  • Shipped production RAG with pgvector/BM25 hybrid retrieval, Reciprocal Rank Fusion, cross-encoder reranking, and LiteLLM routing across Gemini 2.5, GPT-4o, and Claude 3.5; added nightly golden-dataset regression evaluation in CI.
  • Cut monthly hosting costs by ~80% by migrating long-running workloads from Vercel to Fly.io and redesigning the deployment architecture for production.
02
SideQuestSoftware Engineer, AutomationJun 2025 – Dec 2025

Jersey City, NJ

  • Led automation work for a small eng team: shipped internal tools (Instagram data scraper, Google Apps Script jobs) that removed recurring manual data entry.
  • Built an AI quest-generation pipeline (Foursquare Places API, Google Gemini, Unsplash) with batch processing and retries to fetch, enrich, and categorize points of interest.
  • Automated Mapbox tileset generation from PostgreSQL/GeoJSON, powering dynamic location-based quest maps in the product.
03
Montalvo TechnologiesSoftware EngineerJul 2023 – Jan 2024

Hyderabad, TG

  • Built full-stack MERN applications with React interfaces, Node.js/Express APIs, and MongoDB persistence.
  • Implemented CRUD workflows and authentication across frontend and backend application layers.
  • Added API error handling and debugged application behavior across common request failures.
04
TechBullsWeb Development InternJan 2022 – Jul 2022

Hyderabad, TG

  • Integrated a Python stock-price data feed and validated incoming market data for a financial visualization application.
  • Added automated tests for data-processing and calculation logic to verify expected outputs.
  • Built an interactive React/TypeScript interface with Perspective to visualize price movements.

Education

01
Stevens Institute of TechnologyM.S. Computer ScienceDec 2025

Hoboken, NJ

GPA 3.83/4.0

Coursework: Machine Learning, Algorithms, Intro to Cyber Security, Financial Analytics

02
Gandhi Institute of Technology and ManagementB.S. Computer ScienceJun 2023

Hyderabad, TG

GPA 8.29/10.0

Coursework: Data Structures & Algorithms, DBMS, Operating Systems, Computer Networks

OpenSourceContributions

About

Full Stack AI Engineer who builds customer-facing products in React and Node.js and carries features through APIs, data, testing, and deployment. At Fund Flow OS, shipped production RAG, enriched 500+ lead records, and cut hosting costs by ~80%.

Currently
Building a RAG retrieval quality lab
Full Stack AI Engineer at Fund Flow OS
Based in Jersey City, NJ
Certifications

Oracle Cloud Infrastructure 2025 AI Foundations Associate

Oracle · 2025

Skills

Languages & Frameworks

JavaScript/TypeScript · React.js · Node.js · Python · SQL · FastAPI

AI Systems

RAG · Hybrid Retrieval · BM25 · pgvector · Reranking · Vector Embeddings · LLM Evaluation · LiteLLM

Databases & APIs

PostgreSQL · SQLite · REST APIs · GraphQL · OAuth/OIDC

Cloud & Delivery

Fly.io · Vercel · Cloudflare Workers · Docker · Linux · GitHub Actions · CI/CD

Testing & Tools

Pytest · Playwright · Postman · Git

Contact

Let's build

something that

ships.

Open to AI / full-stack roles and founding-engineer conversations.

Email me →GitHub ↗LinkedIn ↗

© 2026 Saketh Kanchi

Built with ❤️