Agentic AI Growth Engineer

Lumida Wealth; Karachi Division; On-site; PKR 33,315,000 - 44,420,000; Full time

Job summary

Lumida Wealth is building an AI-powered a gency ecosystem for creators, owners, and advisors. You will design and deploy agents across customer, prospect, and sales workflows, ensuring compliance and data integrity within a regulated investment advisory framework.

You will integrate HubSpot, Lumida Invest, and experimentation tooling while shaping engagement, outreach, and measurement. This is a hands-on, high-impact engineering role in a US‑Eastern overlap, Karachi-based in-office.

Qualifications

• 3–6 years of engineering experience with production agentic AI systems.
• Python and API fluency with potential to ship production code in Lumida Invest.
• Quantitative depth: experimental design, power and sample size, statistical significance.
• HubSpot and martech experience strongly preferred.
• Content and social automation across blogs, X, LinkedIn, and TikTok with feedback loops.
• Comfort in a regulated financial services environment.
• Strong academic record from a top school.
• Karachi-based, in office, on a US Eastern overlap.

Responsibilities

• Develop AI agents for customers, prospects, and reps.
• Build copilot interfaces for sales reps covering call prep, CRM hygiene, follow‑up drafting, and account research.
• Create a prospect agent to enrich records and score intent with human review.
• Implement a client-lifecycle agent to surface opportunities to advisors.
• Ensure compliance and governance in a regulated environment.
• Integrate HubSpot and Lumida Invest with experimentation workflows.
• Establish dashboards and metrics for sales and CIO desks.

Skills

Engineering experience

Python

APIs

Quantitative methods

HubSpot/Martech

Content automation

Regulated environment

Top school academic record

Education

Strong academic record from a top school

Tools

HubSpot

APIs/tools

Job description

About Lumida

Lumida is an AI-native modern wealth advisor and Wealth Architect for owners and creators.

Our mission is to make money intelligent. We bring investing, trust and estate planning, and tax strategy together in a single wealth architecture—helping clients grow, protect, and transfer wealth with greater intention.

Our investment platform spans public markets and differentiated alternatives including select pre-IPO opportunities for qualified clients. Our approach is institutional in its breadth, non-consensus in its thinking, and deeply personal in its application.

Lumida is a tech-first, AI-native firm built by a technology founder. We are making significant investments in proprietary AI Services that identify opportunities across a client’s financial life, strengthen investment research, automate client‑service workflows, and help our advisors deliver faster, smarter, and more personalized advice.

Our team shares the mindset of our clients: curious, original, and driven to build what does not yet exist. Lumida is venture-backed, with investors including former SEC Chairman Arthur Levitt, and supported by a leadership and advisory network with deep experience across investing, technology, and financial services.

Lumida’s differentiated, non-consensus insights reach a broad audience through our podcast, newsletter, and social media, with recognition across major outlets including The Wall Street Journal and CNBC.

The job in one paragraph

We are building an AI agent for every customer, every prospect, and every sales rep, and the measurement infrastructure to know what works. You will build those agents, wire them into HubSpot and the Lumida Invest app, and run the experiments that decide what we keep. We are increasing marketing spend, which means the app will have the traffic to test properly.

What you will own

Agents for every customer, prospect, and rep.

Four surfaces. A copilot for each sales rep covering call prep, CRM hygiene, follow‑up drafting, and account research. A prospect agent that enriches records, personalizes outreach, and scores intent, with a human sending. A client‑lifecycle agent that watches each client’s situation and surfaces opportunities to their advisor. And a customer‑facing agent inside Lumida Invest.

Building inside a regulated firm.

We are an SEC‑registered investment adviser. An agent that speaks to a client is a regulated surface: what it says falls under marketing and advice rules, and the conversations are subject to retention requirements. Guardrails, disclosure, evaluation, and logging are part of the architecture from the start. Asim owns compliance review and you will work with him closely.

Experimentation infrastructure.

Event instrumentation, assignment and bucketing, an experiment registry, holdouts, and readouts the team can trust. Then running it: A/B and factorial designs on onboarding and the Feed, with power calculations done before the test. The infrastructure needs to be in place and trustworthy by the time the app is at 10,000 daily actives.

Content and social automation.

Pipelines that take our research into blog posts, X, LinkedIn, and TikTok, covering generation, repurposing, scheduling, and performance data feeding back into what gets made next. Our writers own the voice; you own the distribution machine and the feedback loop.

Martech and event triggering.

HubSpot as the system of record, connected to product events, ad platforms, and the agents. Behavioral triggers, lifecycle automation, and closed‑loop attribution from ad click to funded account.

The measurement layer.

Dashboards and metrics the sales team and CIO desk can act on. Every agent and automation ships with an eval and a metric attached.

What we are looking for

• 3–6 years engineering experience, with production agentic AI systems you have built and operated: tool use, orchestration, retrieval, and evals
• Python and API fluency, and the ability to grow into the Lumida Invest codebase and ship production code there over time
• Quantitative depth: experimental design, power and sample size, statistical significance, and judgment about when a result is real. Factorial designs should be familiar
• HubSpot and martech experience strongly preferred — lifecycle automation, event triggers, CRM data models, attribution
• Content and social automation across blogs, X, LinkedIn, and TikTok, understood as systems with feedback loops
• Comfort working in a regulated environment. Financial services experience is a plus, not a requirement
• A strong academic record from a top school
• Karachi-based, in office, on a US Eastern overlap

What success looks like at six months

• The rep copilot in daily use by the sales team
• Prospect research and personalization running automatically, feeding outreach a human sends
• Experimentation infrastructure live, with a documented history of tests, power calculations, and outcomes
• At least one experiment that changed a decision about the Lumida Invest app
• Content and social pipelines publishing across channels with performance data flowing back
• Closed‑loop attribution in HubSpot: a funded account traceable to the ad that started it
• A customer‑facing agent designed and reviewed with compliance, and either shipped or held with a documented reason

Lumida Wealth Management builds for Creators & Owners — founders, operators, and high‑earning professionals. We pay in USD.