Data Scientist

Data Scientist – Core Models (Founding Team)

Affluense.ai | Bengaluru (In-office) | Full-time

Budget for role: 10 - 15 LPA

About Affluense.ai

Affluense.ai is building India's wealth intelligence layer — an AI-powered B2B platform that helps wealth managers, private bankers, and family offices find, understand, and engage High Net Worth Individuals (HNIs) before anyone else does. We aggregate signals from 150+ data sources into 10M+ HNI profiles, and turn that into wealth scores, 360° profiles, relationship maps, and ranked prospect lists that relationship managers actually use to close deals.

We're live with enterprise customers including some of India's largest private wealth and wealth management firms, with a strong active pipeline. We're backed by Zeropearl VC, alongside angel investors Kunal Shah (CRED) and Pravin Jadhav (Dhan).

We're a lean team building something genuinely hard: turning messy, fragmented, often-contradictory data about people's wealth into models that institutions trust enough to base outreach strategy on.

Why this role exists

Right now, the models that power Affluense — wealth scoring, entity resolution across sources, relationship mapping, prospect ranking — are the product. They're also the hardest and most differentiated part of what we do, and the part most in need of a dedicated owner. We're hiring a data scientist to take these from "good enough to sell" to genuinely defensible — the kind of models a competitor can't replicate just by scraping the same public sources we do.

You'll work directly with our CEO (prior data science leadership at HealthifyMe, Mobile Premier League, BrowserStack, and Simpl — where he built alternate-data credit models for India's first BNPL product) and CTO (18+ years in engineering leadership, former Head of Engineering at Edelweiss/Nuvama Wealth). This is a founding-level seat, not a feature-team role bolted onto someone else's roadmap.

What you'll build

- Wealth scoring models — estimating net worth and investable surplus from fragmented, often indirect signals (property records, directorships, public filings, professional signals, behavioral data)

- Entity resolution at scale — matching and deduplicating the same person across 150+ heterogeneous sources with conflicting, incomplete identifiers

- Relationship & network mapping — graph models that surface who's connected to whom, and why that connection matters for an introduction or a deal

- Prospect discovery & ranking — recommendation/ranking systems that tell a relationship manager who to call next, and why

- Feedback loops — closing the loop between real RM usage, deal outcomes, and model retraining/confidence scoring

What we're looking for

- 1-4 years building and shipping ML models in production, ideally on messy, alternative, or adversarial data (credit/risk scoring, fraud detection, identity resolution, recommendation systems)

- Strong fundamentals in classical ML (XGBoost/LightGBM, embeddings) — we're not expecting mastery on day one, just real curiosity and the ability to get there fast

- Solid engineering instincts — you can take a model from notebook to a pipeline that runs reliably, and explain trade-offs in plain language to non-technical stakeholders (including customers)

- Comfort with ambiguity — there's no ground-truth dataset for "true net worth." You'll design proxies, validation strategies, and confidence intervals from first principles

- Bonus: prior exposure to fintech, wealth management, BFSI, or alternate-data credit modeling

Why join now

- You'll be the senior-most — and likely first dedicated — data science hire. Model architecture, tooling, and roadmap are yours to shape, not inherit

- Real customers, real revenue, real usage — your models go live with India's leading wealth management firms, not into a demo

- Direct access to founders with deep data science and BFSI engineering backgrounds