I’m Tanveer. I like turning an idea into something I can see and try, then finding a way to make it better.
This is where I share my personal projects and the stories behind them.
One of the first things I drew on a computer was a park, using keyboard commands in Lotus 1-2-3 on a black-and-white XT.
I wanted to find out whether I could make a scene appear on the screen.
The tools have changed since then. I still enjoy following a question far enough to make something from it.
At university, I added a Pepsi animation to a C++ graphics assignment. Later, breaking my Windows 98 installation gave me a different problem to work through: getting it running again.
Adapted from my LinkedIn story ↗
A memory I wanted to keep. A colour I wanted to get right. A small problem close to home.
Could a game keep a small piece of my daughter’s childhood?
Zara Donut Run carries her original voice recordings into a game she can play on web and mobile. The voice was the reason to make it.
Read Zara’s story
How does a palette feel once it becomes an interface?
Chroa lets me explore colour systems in live interface previews, with accessibility and something reusable to take away.
Explore Chroa
Could finding a halal place to eat feel more like asking someone?
HalalEats SG explores that question through conversation, with certification information sourced from the MUIS Halal Directory.
Try HalalEats SGThere are other experiments too, including Nook, a little home for my coding sessions.
Browse all projectsI’m drawn to the visual side of ideas. Making something helps me understand it. Once I can see it and use it, I start noticing what could be clearer or work better.
I can spend a long time trying different approaches. Friends and family have often reminded me how much creativity is part of that, and I’m learning to give it more room.
A lot of my personal work now involves AI. I share what I’m learning here, including the parts that still need another attempt.
Read my build storiesNow / Sep 2026
Rooms in the work
Code with Claude Singapore, AI Engineer Singapore, OpenClaw Singapore, Singapore FinTech Festival, and more. Lessons from those rooms are published with the builds they changed.
Follow the evidence ↓
Start here
Concepts without theatre, then one small build you can open.
See how a prompt becomes a skill, a harness, and a working product.
Connect experiments to provenance, permissions, and human ownership.
Understand what can stay local - and when cloud models still earn their place.
Curiosity trails
PromptZap opened a question about taste. Chroa made the visual system legible to agents. ScriptToReel carried that thinking into a reproducible media pipeline.
PromptZap → Chroa → ScriptToReelComparing models led to a stronger conclusion: tools, permissions, context, feedback loops, and reusable skills often matter more than a leaderboard position.
Models → coding agents → skillsCloud routing makes comparison easy. Local inference tests privacy, control, latency, and resilience against real workflows instead of slogans.
OpenRouter → Ollama → MLX / oMLXMy fictional Nexus Bank demo is a place to explore permissions, clear stopping points, and when a person should take over.
Personal experiment → Nexus Bank → human reviewA focused Singapore need became a chat-first discovery product grounded in verified public data and the way people actually ask where to makan.
Local context → verified data → HalalEats SGPlay makes feedback immediate. It also exposes when delightful presentation outruns the evidence underneath it.
Voice game → pet league → provenanceThe live lab
Tools, games, prototypes, and the questions behind them. Each project links to something you can try, read, or explore.
A free, local-first screen studio for macOS, built by Tanveer Riaz. Record a display, app, or window, polish the frame and pointer, edit a timeline, and export MP4.
Read about capfr →
A quiet home for existing coding sessions on the CORSAIR XENEON EDGE. Expressive companions, clear states, and a tap back to the work.
Watch Nook and read the story →
A game for my daughter, preserving her little voice from five years ago. Upgraded with Astra from Android to web and mobile.
Play the game ↗Read the story →
A colour-system studio that generates accessible palettes, wires them into live interface previews, and exports a DESIGN.md a coding agent can follow.
Focused, reusable procedures that turn prompt experiments into repeatable execution.
Browse the library ↗
Ask Sherlock Scones for halal places to eat in Singapore, with certification information sourced from the MUIS Halal Directory.
Try the product ↗
A public learning demo for exploring conversational assistance, financial context, and bounded agent actions. It is not a financial service.
Open the demo ↗
Turns rough instructions into structured prompts inside the AI tools people already use.
See PromptZap ↗
A local-first pipeline that researches, scripts, assembles, narrates, and renders short videos from a topic.
Read the source ↗One click turns a webpage into clean Markdown for agents, notes, and knowledge bases. No account or backend.
Install the utility ↗
A playful leaderboard - and a story about replacing persuasive mock data with verifiable public signals.
Read what went wrong →A fast family-friendly experiment in voice input, progressive hints, and playful feedback.
Open on GitHub ↗No public items match this view yet. Try another filter.
Shipping signals
Snapshot supplied on 1 Aug 2026. The total may include private contribution activity; no private repository contents are published here.
Together they are a compact record of fast iteration, collaboration, and pull-request shipping - not a score for its own sake.
Recent credentials · 2025–2026
Tap a card to verify the credential.
AI delivery and data · Issued Jul 2026
Verify ↗ 2025PMIResponsible AI in delivery workflows · Issued Dec 2025
Verify ↗ 2025edXGenerative AI foundations · Issued Apr 2025
Verify ↗ 2025IBF SGAI risk and governance · Issued Mar 2025
View ↗ 2025A-CSMAdvanced agile delivery · Issued Jan 2025
Verify ↗ 2025PMIAI for project leadership · Issued Jan 2025
Verify ↗Learning in public
How tools, memory, permissions, and evaluation turn a model into a dependable workflow.
What can stay on-device without making the product slower or less useful.
How HTML, Markdown, JSON-LD, and clear content policies make a site easier for agents to understand.
A focused, reusable procedure that gives an AI agent domain instructions, constraints, and a reliable workflow for a specific job.
See it in Skillz ↗The tools, context, permissions, memory, feedback loops, and runtime around a model. The harness often decides whether model output can become dependable work.
Read “The harness matters” →AI software designed to keep models, data, or core workflows on a person’s own device whenever practical.
Open the local AI notes →Retrieval-augmented generation: find relevant source material first, then give it to a model as context for an answer.
A record of where data came from, how it changed, and which claims it can legitimately support.
See the leaderboard lesson →Building software through rapid natural-language collaboration with coding agents, with the human retaining intent, taste, verification, and ownership.
Field notes
Build journey · 8 min
Build in public
Build journey
Essay
SG 🇸🇬 Public event trail
01 / 08Singapore FinTech Festival
The conversations left me curious about what happens after a demo works: where the data comes from, what the tool can do, and when someone needs to step in.
A conversation can reveal the missing constraint faster than another hour of polishing. The best notes are usually questions I need to test later.
OpenClaw, Hermes, Sarah, memory, tools and hand-offs all point to the same lesson: autonomy is useful only when its boundaries are visible.
The memorable part was not simply appearing in a group photograph. It was being in a room where practitioners were building, questioning and learning in public.
At a six-hour build table, 87K Windows turned memories from seniors living alone into safe, consented offers of what they could teach. No match is ever faked.
The practical model-cost question is not which model is smartest. It is which model is reliable enough for this job after context, tools, skills, evals and retries are counted.
Model comparison is useful. Reusable skills, clear intent and inspectable review make the learning survive the individual session.
Agent systems become dependable when roles, evidence, review and failure states are designed as one workflow rather than a collection of impressive demos.
We thought about it, planned it, executed and demoed Grub. I got to showcase our dining companion in front of Anthropic.
More rooms, experiments and build notes continue on LinkedIn.
Follow Tanveer's builder journey on LinkedIn ↗Tool relationships
Get oriented broadly, then return to primary evidence.
Different collaborators for different kinds of thinking.
Compare planning, implementation, and review instincts.
Roles, hand-offs, memory, and permission boundaries.
What can stay on-device without forcing every task there.
Visual evidence and story material tied to real builds.
Calmer interfaces for inspecting multi-agent work.
Useful means somebody else can open it - and see its limits.
Active rotation
The models change. The curiosity does not.
OpenAI
Claude
OpenRouter
HHermes
Codex
Cursor
Gemini
Ollama
✦Genspark
Tailscale
>_Ghostty
Railway
Keep the question going
I enjoy seeing how other people make things. If sharing my own attempts helps you try something, I’d be happy with that. This site is for personal projects and community conversations.
Singapore / 2026
The next experiment starts with another question.