Skip to content
Back to roles

Engineering

Applied AI Engineer

Lahore, Pakistan preferred · US consideredHybrid/remotePakistan daytime hours with US overlapFull time

$1,500-$3,500 / mo USD base + equity + profit share

Role at a glance

Owns
AI-assisted workflows, model/prompt choices, structured outputs, evals, and practical automation paths.
Does not own
Research-only prototypes, model demos without product use, or final product prioritization.
Works with
Product, engineering, data, support, operations, and founders.
Success looks like
Manual work becomes faster and more reliable, with quality measured instead of assumed.

About the role

  • Build AI-assisted workflows that make operators faster: lead research, buyer matching, data cleanup, outreach preparation, support triage, and internal automation.
  • Turn repeated manual work into reliable product systems with clear inputs, structured outputs, evals, and human review where needed.
  • Choose models, prompts, tools, and data flows based on measurable business outcomes, not demo appeal.
  • Your objective is to automate the right work after we understand it, simplify it, and know what good output looks like.

What this looks like

  • Operators spend hours researching buyers before outreach. You build an AI-assisted workflow that gathers evidence, produces structured output, and makes uncertainty visible.
  • Support chats contain repeated issues but no one has time to read every thread. You create a triage system that groups patterns, flags customer-impacting friction, and keeps a human review step.
  • A model output looks impressive but is hard to trust. You define evals, failure cases, confidence checks, and review UI so the team can measure quality instead of arguing from vibes.

Where this role stops

  • This is not a demo role. Every AI workflow needs a clear user, failure mode, and quality bar.
  • This is not automation for its own sake. We automate after the workflow is understood and simplified.

Qualifications

  • Strong software engineering fundamentals with hands-on LLM product experience.
  • Comfortable with prompt design, evals, structured outputs, retrieval, and automation workflows.
  • Able to ship practical systems instead of demos.
  • Experience with Python, TypeScript, or both.
Apply now

Compensation

$1,500-$3,500 USD per month as base compensation, based on experience, plus equity and profit-share potential based on seniority, scope, and long-term fit.

Benefits

  • DHA Phase 8 Lahore office with a focused workspace, balcony, clean food, reverse-osmosis water, and on-site office support.
  • Lunch provided for team members working from the Lahore office.
  • Luxury gym membership near the Lahore office for team members who want it.
  • Monthly team activities, optional monthly dinner invites, and annual retreats as the team grows.
  • Generous AI tooling budget for tools that improve your workflow.
  • Near-unlimited Codex access for engineering and product work, with optional Claude or other AI tools by role.
  • Profit-share potential plus equity grants or exit participation for long-term contributors.
  • Clear career ladders in progress, with direct founder feedback and fast ownership paths.
  • Direct access to founders, customers, and real product decisions.
  • High-agency environment with meaningful ownership and fast feedback.

Offer structure

  • Full-time engagement. We don't run part-time trials.
  • 90-day probationary period for mutual fit. Either party can end at any time; payroll for actual days worked.
  • After probation: equity grants, exit participation, or profit-share may unlock based on role, seniority, and long-term fit. 30-day notice on company-side termination.

About Covent

We pivoted, but we are a technology-first real estate company. Being tech first lets us subsidize our costs, so we can make more competitive offers.

It is a very hard business with very slim margins. If it is truly for homeowners, we have to be tech first and cut as many middlemen as possible.

The work matters. If the product is confusing, support gets harder. If the design is sloppy, trust drops. If the code is fragile, the team slows down. We hire people who own the outcome.

Where deals are made

How we work

  1. Question, delete, simplify, accelerate, automate

    We improve work in that order. Ask what problem we are solving, delete what does not help, simplify what remains, shorten the feedback loop, then automate.

  2. Tools are leverage

    We provide the tools, context, and access needed to do ambitious work. The point is not the tool; the point is solving the hardest problem and shipping the clearest system.

  3. Owners, not task completers

    Your job is not just your job description. If you see a better way, say it. If you see a problem, raise it. If you can fix it, fix it.

  4. Speed is learning faster

    Small loops beat big plans. Ask early, ship smaller, surface blockers fast, and communicate with facts: what changed, what is blocked, what is next, who owns it, and when it will be done.

Interview process

Applications are reviewed by founders. No external recruiters. Interviews run virtually or onsite in Lahore. We aim to give a clear yes or no quickly.

  1. Application

    You submit your application. Founders read every one personally. No external recruiters.

    Apply now
  2. Initial interview

    If your application moves forward, we talk. What you have done, how you think about the work, and whether this is actually the role you want.

  3. Project

    A skills evaluation and nothing more. Scoped to what the job actually asks for, so it shows us whether you can do the work rather than testing you for its own sake.

  4. Offer

    Clear the project and you go forward for an offer. We decide quickly, and we tell you either way.

Questions? Email talent@getcovent.com.

Covent is an equal opportunity employer. We hire on merit and fit, not background.