Spotlight Project
NutraLingo
Handles multilingual OCR + real-time ingredient risk scoring in a single PWA pass — the hard part was keeping latency under 2 s on mobile.
Next.jsAIOCR
About
Focused on practical AI products, clean interfaces, and systems that feel usable.
I am Tejas Rawool, a computer science student building AI-driven applications, developer tools, and intelligent systems. My work focuses on turning machine learning ideas into real products that people can actually use.
AI Applications
Machine Learning Systems
Developer Tools
Full Stack Web Apps
Building products where machine intelligence feels clear, fast, and actually usable.
Approach
Start from a real workflow, reduce complexity, then design the ML or AI layer around utility.
Preference
Tools that combine good interface decisions with practical automation and measurable output.
Tech Stack & Capabilities
Engineering Focus & Technical Capabilities
A full-stack foundation evolved into AI engineering — building agentic workflows, typed backend infrastructure, vector search, and production platforms.
Full-stack developer who evolved into an AI engineer building agentic AI systems, backend infrastructure and data platforms.
FRONTEND
Building responsive product interfaces and SaaS dashboards.
ReactNext.jsViteHTMLCSSshadcn/uiTanStack Query
BACKEND & APIs
Building typed APIs, backend services and production workflows.
FastAPINode.jsExpressFlaskREST APIsPydantic
AI / AGENTS
Building agentic workflows, LLM-powered applications and AI-assisted systems.
LangGraphGemini APIGroq APIOllamaLLM IntegrationAI Agents
DATA & DATABASES
Working with transactional data, caching, vector search and persistence layers.
PostgreSQLMongoDBRedisQdrantSQLAlchemyAlembic
AI / ML
Building data-driven applications and machine-learning workflows.
PythonNumPyPandasScikit-learnNLPMachine Learning
DEVOPS & ENGINEERING
Deploying, containerizing and operating modern applications.
DockerDocker ComposeGitGitHubCeleryLinuxNginx
Featured Projects
Featured builds focused on applied intelligence.
A smaller project set does a better job of telling the story. Each build below reflects practical AI thinking, product clarity, and usable engineering decisions.
Project 2
FlowTalk
A real-time multilingual chat platform that removes language barriers through AI translation. People communicate naturally in their preferred language while messages are translated instantly for every other participant.
Zero-latency feel: translation runs server-side on each message event so the sender's experience is unblocked and the recipient sees the result in < 400 ms.
Stack
Next.jsRealtimeTranslation AI
Project 3
AI-Based Resume Screening & Ranking
AI-powered resume screening and ranking system built with Streamlit, NLP, and cosine similarity. Extracts, preprocesses, and ranks PDF resumes against a job description, then presents results in a structured, visualised ranking workflow.
Processes batches of 50+ PDFs in one run and ranks candidates by semantic similarity score — replaces hours of manual screening.
Stack
StreamlitPythonNLP
Project 4
Data Insight Viewer
An interactive, AI-augmented Streamlit app that lets users upload CSV or Excel files and explore datasets through automatic summarisation, visualisations, and lightweight NLP-assisted analysis — no code required.
Auto-generates chart recommendations and natural-language summaries from raw tabular data — designed for non-technical users.
Stack
PythonStreamlitData Visualization
Project 5
GitHub README Globalizer
An Apify Actor that fetches markdown documentation from any GitHub repository, translates it into chosen languages using Lingo.dev AI, and returns commit-ready files while preserving all formatting for developer workflows.
Preserves all markdown structure, frontmatter, and code fences through translation — most tools break formatting on complex READMEs.
Stack
Developer ToolingAILocalization
Project 6
YTDownloadX
A high-performance YouTube video and audio downloader with a polished UI, particle animations, and glass-style effects. Built with Flask and yt-dlp, focused on a fast and seamless media download experience.
Hosted on a self-managed VPS with a custom subdomain — end-to-end ownership from code to deployment.
Stack
PythonFlaskMedia Processing
Career Timeline
Engineering Evolution & Progression
A 4-year journey tracking the transition from core programming fundamentals to machine learning applications, AI products, and production agentic architecture.
1Programming Foundations
2ML + Full-Stack
3AI Products & Automation
4Agentic AI & Systems
Phase 01
Programming Foundations
2023Programming Foundations
Built the Engineering Foundation
Started with Python, JavaScript, C and core programming concepts. Built small projects, practiced problem solving, and learned how software moves from an idea to a working application.
PythonJavaScriptCGit
Phase 02
ML + Full-Stack Applications
2024ML + Full-Stack Applications
From Models to Applications
Moved beyond experimentation into end-to-end applications. Built ML/NLP projects, data-driven tools and web apps while learning how models, APIs, databases and interfaces work together.
MLNLPReactFlaskMongoDBStreamlit
Phase 03
AI Products + Automation
2025AI Products + Automation
Started Shipping AI Products
Shifted from building projects for learning to solving real user problems. Built AI-powered applications, developer tools and multilingual apps like NutraLingo and FlowTalk, securing 2nd Place in the Lingo.dev Hackathon.
Lingo.dev Hackathon — 2nd Place Winner 🏆
AIAPIsAutomationDeveloper ToolsLingo.dev
Phase 04
Agentic AI & Advanced Systems
NOW
2026Agentic AI & Advanced Systems
Building AI Systems & Winning Hackathons
Won 1st Place in the Apify x Lingo.dev Hackathon with GitHub README Globalizer (Multilingual Actor). Currently building production-grade agentic AI systems, multi-agent orchestrations, typed backend services, and scalable data platforms.
Apify x Lingo.dev Hackathon — 1st Place Winner 🥇
LangGraphApify ActorFastAPINext.jsPostgreSQLRedis
PUBLIC WORK & GITHUB
Proof that I build, ship, and maintain.
Projects, repositories, and public development activity that show how ideas move from experimentation to working software.
GITHUB
556PUBLIC CONTRIBUTIONS
SELECTED WORK
4 RepositoriesNutraLingo
LIVE
AI food label scanner with multilingual OCR and real-time ingredient risk scoring.
TypeScript / AI / OCR
FlowTalk
LIVE
Real-time multilingual communication with AI translation per message.
TypeScript / AI / Translation
GitHub README Globalizer
1ST PLACE WINNER
Developer tool for translating GitHub documentation while preserving Markdown structure.
JavaScript / Dev Tooling
FlashChat
OPEN SOURCE
Privacy-focused real-time messaging application with end-to-end user controls.
Full Stack / Privacy / Real-time
GITHUB ACTIVITY
Public development activity — last 12 months
Less
More
TOTAL CONTRIBUTIONS
556Past 1 Year
PUBLIC REPOSITORIES
48Open Source
ACTIVE PROJECTS
18+Shipped & Live
Technical Writing
Writing as a way to explain systems, not just announce them.
The strongest portfolio pieces are backed by technical clarity. These articles show how Tejas turns implementation decisions into readable engineering stories.
LET'S CONNECT
LET'S
BUILD
SOMETHING
MEANINGFUL
Have an idea, project or opportunity?
I'd love to hear and build together.