making ai do more than just yap.
I build AI systems, agents, and automations that handle the tedious stuff while I pretend I planned it that way.
About
i'd rather understand how things work by building them. i spend my time with ml, large language models, ai agents, and full-stack software, turning ideas into working systems and learning a lot from everything that inevitably breaks along the way. lately, i've been really into agentic systems and figuring out how software can become more capable, useful, and a little more autonomous.
Projects
Sellable lets AI agents buy stuff on their own, with guardrails that actually hold. They get a storefront they can read: find products, haggle quotes, pass policy checks, and pay over real Razorpay rails. Every decision lands in an audit ledger. The model suggests, deterministic code decides.
Python · FastAPI · LangGraph · Next.js 16 · Supabase · Razorpay
Helion is an agent harness that lives in your terminal and on your desktop. Watch runs unfold, steer them mid-flight, and approve every tool call from one calm surface. Still cooking, coming soon.
TypeScript · Tauri · Ink · React
Writing
Tech moves stupidly fast, so I write down the things I'm learning, building, breaking, and obsessing over before the next rabbit hole gets me.
Latest
Browse allImplementations
I take interesting papers, ideas, and algorithms and actually build them from scratch.
papers.mahirmalik.inThe Stack
A bunch of languages, frameworks, databases, and AI tools that somehow turn ideas into functioning software. The stack changes, the building doesn't.
Programming Languages
Frontend Engineering
Backend Engineering
Data Analysis & Computing
Machine Learning
AI & LLM Engineering
Agentic Systems
AI Agents
Databases & Data Stores
Workflow & Orchestration
Cloud & Infrastructure
Monitoring & Observability
Developer Tooling
GitHub Activity
A GitHub-style overview of public contributions in 2026, followed by recent public activity and commit messages from @mahirmlk.
Recent Public Activity
Latest visible actions pulled from the GitHub public events feed.
No recent public activity.
At the Moment
Building things, pulling threads, and figuring out why they broke in the first place.
Building
AI systems, agent workflows, and software that does more than sit there waiting for a button.
Learning
LLMs, agent harnesses, retrieval, model tooling, and the engineering around making these systems actually useful.
Looking for
Machine Learning and AI engineering roles, internships, and people working on interesting problems.
Beyond
More systems, deeper work, better agents, and ideas I haven't figured out yet.
