Forward Deployed Data Analyst (Contract-to-Hire)

Fig Labs

is building the

Fig

Analytics Agent—an AI agentic analyst that monitors business data 24 / 7, explains what changed and why, and recommends next-best actions. We help organizations move from static dashboards to

agentic analytics

— systems that think, learn, and act.

We’re hiring a

Forward Deployed Data Analyst

to work directly with customer data, define metrics trees across domains, and implement them on Snowflake, Databricks, and BigQuery. You’ll also codify

agentic analytics

best practices (prompts, patterns, and QA) so our agent delivers deep, trustworthy insights.

What you’ll do

Design domain-specific metric trees

that map business outcomes to drivers and granular KPIs for industries such as :

Sales & Revenue Operations (forecast, pipeline, attainment)

Marketing Analytics (funnel efficiency, attribution, engagement)

Retail Analytics (product availability, store execution, promotions)

Supply Chain Analytics (inventory turns, order fill rate, forecast bias)

Own end-to-end analytics builds

for customers : from source profiling → modeling → metric definitions → validation → visualization → AI-agent prompts.

Implement metrics on modern cloud data platforms

(Snowflake, Databricks, BigQuery) using

SQL

(and dbt or equivalent) with tested, reusable patterns.

Author agentic analytics prompts

for the Fig Analytics Agent :

Few-shot exemplars, tool-use patterns, system / task prompts, safety / guardrails, evaluation criteria, and correction loops.

Create visualization playbooks

(chart choices, small-multiples, uncertainty bands, baseline / target overlays) to communicate insights clearly.

Partner with stakeholders

(RevOps, Marketing Ops, Merch / Ops, Supply Chain, Finance) to translate questions into analyzable hypotheses and decision-ready outputs.

Harden quality : write QA queries, metric unit tests, anomaly checks, and reconciliation steps (source vs. modeled vs. dashboard).

You’ll be great at this if you have

Expert SQL

(window functions, CTEs, semi-structured data, performance tuning) and strong experience on

Snowflake, BigQuery and / or

Databricks.

Experience with RevOps, marketing analytics, retail analytics, or supply-chain analytics (consulting or forward-deployed roles a plus)

Hands-on with

data modeling

(star schemas, facts / dimensions, SCDs), semantic layers, and

dbt

(or similar) best practices.

Proven

visualization sense

(knowing when to use bars / lines, cohort heatmaps, waterfalls, funnel and Sankey for flow, control charts for stability, prediction intervals).

Comfortable building and validating metrics for :

Funnel analysis

(MQL→SQL→Opportunity→Win)

Forecasting & reconciliation

(weighted pipeline, quota / target alignment, MAPE / WAPE)

Cohort & retention ,

RFM / segmentation

Market Basket Analysis / affinity ,

ABC / XYZ classification

Attribution

(rules-based, simple MMM / MTA concepts)

A / B testing

basics and lift analysis

Anomaly detection

& outlier triage

Familiarity with

agentic / LLM workflows

(prompt design, few-shot patterns, retrieval / context packaging, evaluation / guardrails) or strong motivation to learn fast.

Excellent stakeholder communication—turn ambiguous questions into crisp hypotheses, datasets, and decisions.

What success looks like (90 days)

30 days :

Current customer schema profiled; baseline

metrics tree

drafted for one domain (e.g., Sales) and validated with stakeholders.

60 days :

Metrics implemented on a target platform (Snowflake / Databricks / BigQuery) with QA checks; first

agentic prompts

producing reliable, cite-able analyses.

90 days : Reusable

agentic analytics playbook

(prompts + SQL patterns + tests + viz guide) shipped; two domains live (e.g., Sales + Marketing) with measurable decision impact.

Why Join

Be part of a team shaping the future of

autonomous analytics agents .

Work directly with the founder team on

real customer datasets and live demos .

Ship high-impact analytical content seen by enterprise audiences.

100% remote and flexible — output and curiosity matter more than location.