Building AI products
that people actually use.
I'm Lovepreet Singh, an AI-focused developer who enjoys transforming ideas into real products using modern web technologies, APIs, and artificial intelligence.
Products I've shipped, not just demos.
A selection of AI products and tools I've built and deployed. Each one is live and serving real users — no vaporware.
ToS Summarizer
Summarizes Terms of Service using AI
Paste any Terms of Service document and get a plain-English breakdown of the clauses that actually matter — data sharing, arbitration, cancellation, liability.
More on the way
Reserved for the next idea
I ship faster than I document. This slot is reserved for the next product — likely something at the intersection of agents, retrieval, and a real workflow people hate doing manually.
More projects on GitHub
Smaller experiments, demos, and work-in-progress live in my repositories.
A place where I experiment with machine learning, language models, APIs, and new product ideas.
Not every experiment becomes a product. Some are research, some are toys, some teach me something I'll use six months from now. All of them are tracked here.
AI Clone
Fine-tuned model that mirrors my writing style, deployed behind a clean inference endpoint.
Voice Model
Experimenting with voice cloning for short-form narration and accessibility use cases.
Prompt Engineering
A library of reusable, version-controlled prompt templates evaluated against real tasks.
Image Enhancement
Diffusion-based upscaling and restoration pipeline powering EnhanceIt.
RAG Experiments
Comparing retrieval strategies, chunking, and reranking for grounded LLM answers.
Future Ideas
Agent workflows, on-device inference, and a few products I'm not ready to talk about yet.
A short version of a longer story.
The arc, in four points and four milestones.
Interest in AI
I treat AI as a building material, not a buzzword. Models are tools — the interesting work is wiring them into products that earn a spot in someone's daily workflow.
Building products
I ship. A working product beats a perfect prototype. I'd rather launch something rough and iterate against real usage than polish in private forever.
Learning continuously
The field moves weekly. I keep a steady cadence of reading, building, and breaking things — most of what I know came from finishing small projects, not courses.
Solving real problems
I look for problems people already pay for, hate doing, or do badly. AI is at its best when it removes drudgery from a workflow that already exists.
Timeline
Started programming
Picked up Python out of curiosity, fell in love with the feeling of turning a blank file into something that actually does something.
Built first public project
Released StreamPoint as my first publicly available product. Learned the difference between code that runs on my machine and a product real people can use.
Released AI applications
Shipped EnhanceIt and the Personal AI Clone. Started treating models as building blocks for products rather than academic exercises.
Exploring AI agents & ML
Currently going deep on autonomous agents, retrieval pipelines, and on-device inference — building toward a product that does real work, not just demos.
Tools I reach for first.
The stack changes as the field does. These are the tools I've shipped with recently — grouped by what they're actually for.
Activity on GitHub.
A snapshot of recent repositories and a placeholder contribution graph. The real one loads from the GitHub API in production.
Contribution activity
Last 12 months · 1,248 contributions
Have an idea worth building?
I'm currently open to AI product work, collaborations, and the occasional interesting side-project. Drop me a line.