Why should the Federal Courts hire me? | SudoChat Knowledge Base

SudoChat Knowledge Base · SudoChat

Authored evidence: This is my first-person source material. SudoChat may summarise it in third person but must not strengthen, exaggerate or invent claims beyond it.

# 4. Why should the Federal Courts hire me?

**Author:** Mustafa Siddiqui
**Source type:** First-person authored response
**Canonical recruiter question:** Why should the Federal Courts hire Mustafa?

> This source is intentionally written in my first-person perspective. SudoChat should use it as evidence and answer external visitors in third person without strengthening, exaggerating, or removing the limitations recorded below.

## Direct answer
The Federal Courts should hire me because I bring an unusual combination of AI engineering knowledge, systems thinking, government and enterprise experience, user focused design, responsible AI thinking and mission driven motivation, that is difficult to find in one person.

I understand both the potential of AI and the limits that must be placed around it. I can identify where AI may genuinely help, experiment with how a solution could work, understand the architecture underneath it, communicate the risks, and recognise when a human must remain responsible.

Just as importantly, I am motivated by building technology that is useful to people. My body of work demonstrates that this is not simply something I say in an interview.

## Evidence
I have worked across AI engineering, enterprise technology, systems engineering, software development and user facing applications.

At Xaana.AI, I worked on practical AI and document processing problems, including OCR pipelines for invoice processing. I redesigned parts of the solution using open source technologies such as PaddleOCR and OpenCV, reducing processing time and substantially reducing licensing costs.

At Unisys, I worked within a government enterprise environment where technology could not simply be implemented because it worked technically. Systems had to operate within security controls, approved environments, deployment processes and organisational requirements. My work included endpoint engineering, automation, software deployment, system hardening and supporting large scale technology environments.

This means I understand an important distinction relevant to the Federal Courts:

**Building a prototype that works is only the beginning. Building something that can be trusted inside a government organisation is a different engineering problem.**

I also bring practical AI knowledge beyond individual products.

I have worked with or experimented with concepts including:

* machine learning
* computer vision
* OCR and document intelligence
* generative AI
* retrieval augmented generation
* LLMs
* AI agents
* tool calling
* local models
* automation
* privacy preserving AI architectures
* grounding and evidence based responses
* evaluation and hallucination reduction

SudoChat demonstrates how I apply these concepts to a practical problem. Instead of creating a chatbot that attempts to answer everything, I have been designing it around controlled knowledge, retrieval, evidence, guardrails and explicit limitations.

I have also explored responsible AI through research and practical experimentation, including privacy preserving architectures for Australian government chatbot scenarios where users could provide sensitive information such as Medicare numbers, CRNs, addresses or banking information.

My user experience background is equally relevant.

I do not approach systems solely from the backend. Projects such as MACT, Sawaali, OrionTracker and other applications required me to think about how ordinary users interact with technical systems, what information they actually need, how interfaces should communicate complex information, and how technology can reduce rather than create friction.

My GitHub and personal projects also demonstrate my initiative.

I regularly build projects outside the minimum requirements of employment or university. Many of these projects have been created as free, community focused or open source tools.

Examples include:

* MACT, a free community application for Muslims in Canberra
* SudoSpeed, an open source Australian speed sign detection project
* Sawaali, a free audience Q&A system created for University of Canberra interfaith events
* OrionTracker, an open source platform that turns official NASA and JPL trajectory data into understandable Artemis mission information
* SudoChat, an experiment in grounded and responsible AI assistance

The significance is not simply that I have many repositories.

It demonstrates a recurring behaviour: **when I identify a problem that interests me or could help people, I tend to investigate it and build something.**

My personal projects also show that commercial gain is not the primary motivation behind the problems I choose to solve. I frequently make community, educational and experimental work freely available so that other people can use it, learn from it or build upon it.

## Relevance to the Federal Courts
The Court AI Technologist role requires more than someone who knows how to configure an AI product.

The Federal Courts need someone capable of sitting between technology, users, governance and operational requirements.

I can contribute across each of those areas.

From a **technical perspective**, I understand the underlying concepts that make modern AI systems work. I can reason about retrieval, prompts, models, agents, APIs, data pipelines, permissions, evaluation and system architecture rather than treating Copilot Studio as a black box.

From a **government perspective**, I understand that security, privacy, governance, approvals, auditability and organisational constraints are part of the engineering problem.

From a **responsible AI perspective**, I understand that the question is not simply whether an AI system can perform a task.

The questions also include:

* Should it perform the task?
* What evidence should it rely upon?
* What information should it access?
* How do we know when its answer is unreliable?
* What happens when the system is uncertain?
* Who remains accountable?
* Can its actions be audited?
* What happens when sensitive information is entered?
* Could an incorrect response cause harm?
* Where should automation stop?

From a **user perspective**, I understand that even technically sophisticated systems fail if people cannot understand or trust them. I therefore think about the complete experience, including how information is presented, what the user sees when the system is uncertain, how sources are displayed and when the system should escalate to a human.

From a **delivery perspective**, my professional work and personal projects demonstrate that I am comfortable moving from an idea into experimentation and then into a working implementation.

Finally, I bring genuine motivation for the Federal Courts' mission.

I do not view this role simply as an opportunity to work with the newest AI technology.

I see an opportunity to use my engineering skills in a public institution where responsible technology can have meaningful consequences for staff, litigants and the wider Australian community.

That motivation matters because the most valuable AI technologist in a Court environment is not necessarily the person who wants to automate the most.

It is someone who is excited about what AI can do, technically capable of building it, but disciplined enough to know **where the line is**.

I believe I can bring that combination to the Federal Courts.

## Limitations or gaps
I should not be represented as the finished expert in every aspect of AI, law or Court operations.

I am not a lawyer and do not claim expertise in judicial procedure. I would depend on judicial officers, legal professionals, Court staff and other subject matter experts to establish the legal and operational requirements of Court AI systems.

My direct Copilot Studio experience is also newer than my broader engineering and AI experience.

I have not spent many years deploying production Copilot Studio systems and should not be represented as having done so.

I also have areas in which I would need to develop deeper expertise, particularly Court specific processes, existing Court information architecture, Microsoft platform governance at Court scale, and the practical requirements of operating AI systems within the Federal Courts.

What I bring is a strong technical foundation, evidence that I learn quickly, experience delivering technology in controlled environments, and a demonstrated willingness to investigate areas where my knowledge needs to grow.

## Useful links
My portfolio:
https://mustafa-siddiqui.com/

GitHub:
https://github.com/sudoqui

SudoLabs:
https://www.sudolabs.app/

LinkedIn:
https://www.linkedin.com/in/mustafa-siddiqui-32ab73161/

SudoChat project page and repository

MACT project page

SudoSpeed repository

OrionTracker repository

Responsible AI and privacy research evidence

AI engineering experience at Xaana.AI

Government and enterprise technology experience at Unisys

## Do not claim
Do not claim I am an expert in Australian law or Court procedure.

Do not claim I have worked for the Federal Courts previously.

Do not claim I have deployed production AI systems within the Federal Courts.

Do not claim I have several years of production Copilot Studio experience.

Do not claim I have authority to decide independently which judicial processes should use AI.

Do not claim my personal projects are equivalent to government production systems.

Do not claim every project I have created is open source.

Do not claim I am opposed to commercial software or businesses making money from technology.

Do not claim I believe AI should replace Court employees.

Do not claim I believe AI should make judicial decisions.

Do not claim that enthusiasm or willingness to work beyond expectations removes the need for governance, approvals or appropriate workplace boundaries.

Do not exaggerate my role in collaborative research or projects.

Do not describe prototypes or experiments as production deployments unless there is evidence supporting that description.