A question I keep returning to: how can software expand human capability without erasing judgment, context, or accountability?
I started in information security and ethical hacking. Later, my work moved into growth marketing, enterprise digital operations, and AI transformation. The job titles changed, but the habit stayed the same: understand how the system behaves, make the complexity usable, and build something people can actually rely on.
Most of what I’m exploring now sits where AI systems, marketing engineering, memory, identity, and human judgment meet. I prefer to build before I explain. The repositories below are the parts you can inspect, run, question, and improve.
Outside software, I write and paint. I think in Bengali as often as I do in English. For me, technology becomes useful when it meets a real person, a real organization, and a real consequence.
| Layer | The question it answers |
|---|---|
| Context & memory | What should the system know, retain, and learn from? |
| Identity & authority | Who—or what—is acting, and with whose permission? |
| Agents & workflows | How does useful work move from intent to durable output? |
| Evidence & evaluation | What supports the answer, and how do we know it holds up? |
| Human approval | Which consequential decisions must remain accountable to a person? |
The standard I care about is simple: show the evidence, make the limits visible, and don’t confuse a polished demo with a dependable system.
An open-source marketing operating system that carries brand context through strategy, SEO, AEO/GEO, paid media, content, CRM, and analytics—with human approval and verifiable outputs built in. |
An early-stage architecture for agent-native marketing operations: capabilities instead of simulated departments, explicit autonomy controls, layered context, memory, and accountable approval. |
An editorial production system built around research, claim verification, review gates, source-aware internal linking, publication-ready delivery, and provenance. |
A social production system that turns a content calendar into on-brand copy, images, video, and client-ready delivery—with compliance checks and human approval at consequential steps. |
Choose your route: run the broader marketing system · study the agent architecture · build long-form content · build social content · install the suite
Earlier and experimental public work
| Project | Place in the story |
|---|---|
| AEO Strategist | An early open-source skill for turning user-supplied research into structured answer-engine strategy. |
| OpenFire AdminFinder | A 2013 Perl utility from my information-security years, preserved as part of the back catalogue. |
| OpenFire Technologies | The archived public footprint of that earlier security chapter. |
| All public repositories | Coursework, experiments, forks, current systems, and the rest of the visible record. |
| Agent-native organizations How should context, authority, capabilities, memory, and accountability fit together? |
Persistent knowledge & personal AI What should an AI remember—and how should that memory earn trust? |
| Digital identity & delegation How do we make it clear who is acting, with whose authority, and who remains responsible? |
Human capability How can software extend judgment and agency without flattening either? |
These are questions I’m actively exploring, not finished products.
I write from the work: AI systems, marketing engineering, digital operations, and the practical edges of human–AI collaboration.
| Read | Watch & follow | Discuss |
|---|---|---|
| indranil.in Essays, notes, and the wider body of work. |
YouTube · LinkedIn · X Videos, shorter-form thinking, and ongoing conversations as @askneelnow. |
GitHub Discussions Questions and ideas around the public systems. |
