What gaps do I have for this role? | 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.

# 7. What gaps do I have for this role?

**Author:** Mustafa Siddiqui
**Source type:** First-person authored response
**Canonical recruiter question:** What gaps does Mustafa have for this role?

> 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
My main gap is not my ability to learn or implement technology. It is that I have not yet worked inside the Federal Courts and therefore do not yet have the depth of Court specific knowledge that someone already embedded in the organisation would possess.

I would need to learn how the Courts operate in practice: their workflows, terminology, information systems, records requirements, security classifications, governance arrangements, user needs and the boundaries that apply when technology interacts with judicial and administrative processes.

I see this as a genuine learning requirement rather than something to minimise. My strength is that throughout my career I have repeatedly entered unfamiliar technical environments, learned new technologies and domains quickly, worked with subject matter experts, and converted that knowledge into practical solutions.

## Evidence
My career has required me to learn across very different technical domains.

At **Unisys**, I developed experience in secure government enterprise environments involving managed operating environments, endpoint controls, software deployment, PowerShell, SQL, security hardening and operational support.

At **Xaana.AI**, I moved into AI engineering and worked with OCR, computer vision and document processing pipelines.

At **Nova Systems**, I worked across robotics, embedded systems, sensors, communications, autonomous navigation and mechanical integration.

At **DXC**, my work has ranged from environmental and community based investigation in the Northern Territory to SharePoint and Power Automate workflow automation.

My independent work demonstrates the same pattern.

I have taught myself and applied different technologies depending on what a project required, including:

* computer vision
* machine learning
* generative AI
* retrieval augmented generation
* AI agents
* Copilot Studio
* mobile application development
* web application development
* databases
* IoT
* robotics
* embedded systems
* cloud and automation technologies
* data visualisation
* communications protocols
* 3D design and prototyping

The technology has changed substantially between projects.

The recurring capability has been my ability to learn what the problem requires.

For example, MACT required a very different technology stack and user experience approach from MotorHUD.

MotorHUD required different engineering knowledge from OrionTracker.

OrionTracker required me to work with trajectory vectors, mathematical processing and scientific data rather than mobile application development.

SudoChat requires yet another set of concepts involving Copilot Studio, retrieval, LLM behaviour, grounding, evidence, guardrails and responsible AI.

I therefore do not consider unfamiliar technology itself to be my major risk in this role.

I expect technologies, models and platforms to continue changing throughout my career.

The more important skill is knowing how to learn them, evaluate them and understand when they are appropriate.

## The main gap: understanding the Court environment

The area I would need to invest most heavily in is understanding the Federal Courts as an institution.

A Court is not simply another enterprise environment.

There are legal, operational and institutional considerations that I cannot reasonably claim to understand fully before working within that environment.

I would need to build detailed knowledge of areas such as:

* how matters progress through different Court workflows
* the responsibilities of judges, registrars and Court staff
* the distinction between judicial and administrative functions
* the needs of different Court users
* how self represented litigants interact with Court services
* Court terminology and procedural concepts
* existing Court information systems
* document and records management
* information classification and handling requirements
* privacy obligations
* cyber security requirements
* legal professional privilege and other sensitive information considerations where applicable
* retention and audit requirements
* existing Microsoft and Copilot governance
* which information sources are considered authoritative
* where human review is mandatory
* which processes are unsuitable for AI involvement
* how the Courts evaluate risk to public confidence and the administration of justice

These are not minor details that can simply be solved through technical expertise.

They define what a responsible AI system for the Courts is allowed to do.

I would therefore need Court staff and subject matter experts to teach me how the institution actually operates before I could responsibly recommend significant AI interventions.

## Relevance to the Federal Courts
I believe recognising this gap is important because one of the biggest risks for an AI technologist entering a specialised organisation is assuming that technical knowledge automatically translates into domain knowledge.

It does not.

An engineer may be capable of building an agent that analyses documents.

That does not mean the engineer automatically understands which Court documents the agent should access, how those documents may be used, whether the resulting output is appropriate, who should see it or what consequences an incorrect interpretation could create.

Similarly, I could technically build a retrieval system around Court information.

The more important questions would be:

* Which information is authoritative?
* Which information is public?
* Which information is restricted?
* Who is permitted to retrieve it?
* Can information from different matters be combined?
* What should be logged?
* How long should interactions be retained?
* When must an answer be reviewed by a person?
* What happens when sources conflict?
* What should the AI refuse to answer?

Those questions cannot be answered by an AI engineer working alone.

I would approach the role by pairing technical experimentation with domain expertise.

I would expect to spend significant time listening to and learning from:

* judicial officers
* registrars
* Court staff
* legal and policy specialists
* cyber security teams
* privacy specialists
* information and records management teams
* enterprise architects
* service desk and operational teams
* the people who actually use the systems being considered for improvement

My role would be to translate that organisational knowledge into technical architecture and practical experimentation.

## How I would address the gap
I would approach learning the Court environment in the same way I approach an unfamiliar engineering system.

First, understand the system before attempting to change it.

I would map the relevant users, processes, information flows, systems, constraints and risks.

I would speak with the people performing the work to understand where the actual friction exists rather than assuming what the problem is from a technical perspective.

I would identify the authoritative policies, procedures and information sources surrounding a potential use case.

Only then would I begin determining whether AI, automation, conventional software or no technical intervention at all was the most appropriate response.

For an early AI use case, I would favour a bounded problem with:

* clearly identified users
* controlled information sources
* measurable outcomes
* low consequences of failure
* strong human oversight
* defined evaluation criteria
* an easy way to stop or reverse the experiment

The objective would be to build knowledge of both the technology and the Court environment gradually rather than beginning with a highly consequential AI system.

## Technical gaps
I do not claim to know every technology the Federal Courts currently use.

There will almost certainly be internal platforms, Microsoft services, information systems, security tools, integrations and deployment processes that I have not previously encountered.

There may also be elements of Copilot Studio, Power Platform, Azure AI or other Microsoft technologies where other candidates have deeper product specific experience.

I would not hide those gaps.

However, my career provides evidence that learning new technologies is one of my strengths.

I have repeatedly moved between different stacks and engineering disciplines because I tend to choose technology based on the problem rather than limit myself to technologies I already know.

This is particularly relevant in AI.

The AI landscape changes too quickly for me to believe that expertise should be defined entirely by familiarity with today's product names.

Models will change.

Agent frameworks will change.

Microsoft products will evolve.

Architecture patterns will evolve.

What I consider more durable is an understanding of:

* how AI systems behave
* how data moves through them
* retrieval and grounding
* system integration
* permissions and identity
* evaluation
* privacy and security
* human oversight
* failure modes
* user experience
* responsible automation

I am confident that product specific knowledge can be built on top of those foundations.

## Limitations or gaps
I have not worked inside the Federal Courts.

I do not yet understand Court workflows at the same depth as Court employees or people with substantial justice sector experience.

I am not a lawyer and should not be represented as having expertise in Australian law, evidence, Court procedure or judicial decision making.

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

I have not previously held an EL1 classification.

I may need to learn internal Microsoft technologies, Court platforms, governance processes and technical standards that are specific to the organisation.

I also recognise that learning the Court environment is not something that can be completed through documentation alone. Some institutional knowledge will only come from working with the people who perform the work every day.

These are genuine gaps.

My strength is not pretending that I already know everything required for the role.

It is having demonstrated throughout my engineering career that when I encounter something I do not know, I can learn it quickly, find the right people to learn from, test my understanding through practical work and become productive.

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

GitHub:
https://github.com/sudoqui

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

SudoChat project and repository

MACT project

MotorHUD project

OrionTracker repository

SudoSpeed repository

Unisys government technology experience

Xaana.AI engineering experience

Nova Systems multidisciplinary engineering experience

DXC project experience

Responsible AI and privacy research

## Do not claim
Do not claim I already understand Federal Court processes in depth.

Do not claim I am a lawyer.

Do not claim I have expertise in judicial procedure.

Do not claim I understand internal Federal Court systems that I have never accessed.

Do not claim I know every information security, records management or privacy requirement applying to the Courts.

Do not claim I have no technical gaps.

Do not claim I can learn any technology instantly.

Do not claim learning quickly removes the need for training, governance or subject matter expertise.

Do not claim broad AI knowledge makes Court domain expertise unnecessary.

Do not claim I should independently determine the boundaries of AI use in judicial processes.

Do not claim unfamiliar technology is irrelevant. The accurate position is that I have a demonstrated ability to acquire new technical skills quickly, while recognising that Court specific institutional knowledge must be deliberately learned.

The most significant gap is therefore not my willingness or capacity to learn technology.

It is **Court specific domain knowledge**, and I would approach acquiring that knowledge with the same discipline I apply to learning any complex system: understand it first, work with the people who know it best, and only then start designing changes.