Senior Data Engineer

Starting Salary: Rs 501,000 PKR / month

Schedule: U.S. Eastern, Monday through Friday

Location: 100% Remote, Full-Time

About the Role

Falcon Scaling and PolarityIQ build private capital markets intelligence products. Our clients are family offices, institutional LPs, fund managers, investment banks, and registered investment advisors. They pay premium prices for verified, decision-maker-level intelligence on capital allocators worldwide.

We operate a commercially active library of more than two hundred data products spanning family offices, institutional investors, venture capital, private equity, fund of funds, mutual funds, real estate, and high net worth individuals. Our flagship products include Family Office MAX (25,000+ executives across 10,000+ family offices) and LP MAX Bundle (80,000+ contacts across 17,000+ institutional LPs representing over $2 trillion in AUM).

The Senior Data Engineer (Full Chain Builder) owns Project NOVA: the three-stage production pipeline that acquires records at scale, enriches them to the granularity capital allocators will pay for, and activates them inside PolarityIQ through agentic retrieval that surfaces signals autonomously.

This is hands-on, high-autonomy work. You will build original scraping pipelines, architect multi-provider enrichment waterfalls, deploy production RAG systems, and build the agentic retrieval layer across our intelligence library. You ship products paying customers rely on, not internal demos.

Before You Apply

If your resume does not name a specific client, a real user volume, and what breaks when your system goes down: do not apply.

If you have not deployed at least one production RAG system that a non-technical user relied on: do not apply.

If Claude Code is not already part of your daily workflow, this role will move too fast for you.

Compensation

Starting salary is Rs 501,000 PKR per month. The role includes a revenue-sharing component tied directly to your contributions. You are paid weekly, on time.

Strong performers regularly see total earnings scale meaningfully within the first four to eight weeks, with continued upside as contributions compound. On-target monthly earnings can exceed Rs 1,000,000 PKR within six to eight months.

How You Enter — The 30-Day Sprint

This role begins with a paid, deliverable-based 30-Day Build and Monetization Sprint at Rs 501,000 per month. You do not need to leave your current position. You operate on your own schedule, deploy existing products against real demand, and report what happens when your work meets the market.

This is not a trial period. It is a real operating environment. At day 31, both sides hold 30 days of evidence. You will know whether the earnings trajectory and autonomy are real. We will know how you operate under real conditions.

What You'll OwnAcquire — Sourcing and Scraping

• Build scraping pipelines against family office sites, regulatory filings, SEC EDGAR, state corporate registration databases, and professional networks.
• Handle production anti-bot environments — Cloudflare, DataDome, PerimeterX — through residential proxies, CAPTCHA solvers, and stealth browser configurations.
• Orchestrate recurring scraping with Apache Airflow, Prefect, or equivalent. Scheduled, monitored, recoverable.
• Validate sourced records against schema contracts using Pydantic or equivalent. Every record passes the validation gate before entering the pipeline.

Enrich — Multi-Provider Waterfalls

• Build and manage multi-source enrichment waterfalls using Clay.com, Apollo, Hunter, People Data Labs, Clearbit, and similar platforms.
• Know the cost, coverage, and reliability tradeoffs between providers field by field. Build waterfalls that minimize cost while maximizing coverage.
• Add granularity dimensions that matter: decision-maker layer, investment thesis, activity signals, relationship graph, structural classification, validation metadata.
• Apply data hygiene principles across all pipelines: bounce rates, email verification, contact scoring, stale record detection.
• Operate orchestration tooling that lets enrichment chains run end-to-end without manual stitching: n8n, Apache Airflow, Prefect, Make, or equivalent. The waterfall is one workflow, not a sequence of disconnected scripts.
• Build cost-efficient pipelines that scale without proportional cost growth.

Activate — Agentic RAG and Retrieval

PolarityIQ's core intelligence layer is built on RAG. This is not experimental. It is the foundation of how our platform transforms raw family office data into queryable, actionable intelligence.

• Architect end-to-end RAG systems: document ingestion, chunking strategy, embedding model selection, vector database architecture, retrieval pipeline design.
• Select and justify vector database choices (ChromaDB, Pinecone, Supabase pgvector, Weaviate, or equivalent) based on cost, latency, and scale requirements.
• Implement hybrid retrieval strategies combining semantic vector search with keyword matching (BM25) for higher relevance and accuracy.
• Build retrieval evaluation loops. Assess whether your system returns accurate, grounded results and diagnose when it does not.
• Design and deploy autonomous AI agents using LangChain, CrewAI, or equivalent frameworks. Agents that gather, validate, and synthesize intelligence across data sources.
• Integrate agentic workflows into the PolarityIQ platform so users receive signal delivery, not just search results.

Database Architecture

• Own the architecture, integrity, and growth of PolarityIQ's core intelligence database.
• Design and manage relational schema across Supabase/PostgreSQL, ensuring normalization, indexing, and query performance at scale.

Technical StackScraping and Acquisition

Scrapy, Playwright, curl_cffi, BeautifulSoup, Requests. Apify, Phantombuster, Browse AI at production scale. Anti-bot fluency: Cloudflare, DataDome, PerimeterX (named experience, not theoretical). Proxies: Bright Data, Oxylabs, Zyte, Smartproxy. CAPTCHA solvers: 2Captcha, Anti-Captcha, CapSolver. Orchestration: Apache Airflow, Prefect, AWS Lambda.

Enrichment

Clay.com as primary environment: tables, waterfalls, formulas, AI columns. Provider depth: Apollo, Hunter, People Data Labs, Clearbit, BetterContact, Cognism. Validation: Pydantic, Pandera, or Great Expectations applied to scraping output.

RAG and Agentic Systems

Production RAG deployment covering document ingestion, chunking, embeddings, hybrid retrieval, re-ranking, grounding. Vector databases: ChromaDB, Pinecone, Supabase pgvector, Weaviate. Agent frameworks: LangChain, CrewAI, LlamaIndex. LLM APIs: OpenAI, Anthropic Claude, Gemini, with understanding of cost, latency, and reliability tradeoffs.

Claude Code — Non-Negotiable

You operate Claude Code in terminal-based, agentic workflows. Not as a code completer but as a development collaborator you direct, correct, and build with autonomously. You know its limits and call them without hand-holding.

Development and Infrastructure

Backend with Python, Node.js, or similar. SQL, relational schema design, query optimization. Managed platforms (Supabase, PostgreSQL). API integration (REST, GraphQL, webhooks). Git / GitHub.

Who You Are

The technical skills matter. This section matters more.

You take ownership before being asked. You measure yourself by outcomes, not activity. You are honest, even when it is uncomfortable. You are reliable, not occasionally but consistently. You are resourceful enough to treat every dollar spent like it is coming out of your own pocket.

Your thought process is disciplined. You verify before you assert. You distinguish between facts, inferences, and conclusions, and you never conflate them. You test your assumptions before acting. You are willing to challenge your own hypotheses when the evidence points elsewhere.

You use AI extensively but think independently first. Claude Code feels like a natural extension of how you work. You know when it is wrong, when it is shallow, and when to apply the human layer of judgment no prompt can replicate.

You communicate in flawless English. Your first draft should be publishable. You do not require heavy editing cycles.

Job Type: Full-time

Pay: Rs501,000.00 per month

Education:

• Master's (Preferred)

Language:

• English (Preferred)

Work Location: In person