Awais faces a mountain ridge at sunset

Awais Arif

Forward Deployed AI Engineer

I turn manual research and review into software your team can check.

View Projects

Experience

Research tooling and full-stack delivery.

  1. CodeBricks Global

    AI/ML Engineer

    I turn consumer research methods into software analysts can run, including survey classification, validation, and statistical modelling.

  2. Gigalabs

    Software Engineering Intern

    I was the sole developer of an internal platform used by more than 50 employees. I also contributed to the Plexxis enterprise project.

  3. Independent Client Work

    Freelance Software Engineer

    For a medical consultancy, I built the site, set up the domain, and gave patients a way to book appointments online.

Selected Projects

How I Work

I work with the people who do the job today. They help me spot the exceptions and check the first version against their own work.

I studied computer science at COMSATS. Outside engineering, I take my camera into the mountains.

Awais Arif in a dark jacket and cap with mountains behind him
  1. Start With a Real Example

    Walk through a survey response, a document question, or a record your team reviews.

  2. Agree on a Good Result

    Pick examples to test and decide which cases still need someone’s judgment.

  3. Prototype, Pilot, Release

    Build one useful part first. Test it on your examples before connecting it to the rest of the workflow.

Frequently Asked Questions

What kinds of projects do you take on?

I build tools for teams that spend hours going through feedback, finding answers in documents, or checking records against each other. A good starting point is a job someone on your team already does and can show me.

What should I bring to the discovery call?

Bring one example of the work, such as a survey response, a document question, or a record someone has to review. Tell me what takes the most time and what a good result looks like. An anonymized example works well.

Can you work with the tools we already use?

We’ll start with where the work happens today. A first version can often use an export from your existing tools. If that saves time, we can look at connecting the systems directly.

How do you check the AI’s output?

I test it against examples your team has reviewed, including cases people disagree on. Depending on the task, I’ll show the source behind an answer, let reviewers correct a tag, or send uncertain results to a person.

What Does Your Team
Still Do by Hand?

Bring one workflow and an example you deal with today. We’ll map where the time goes and what a useful first version could handle.

Book a Discovery Call

Choose a time for a Google Meet call.