Lead Analytics Architect – Design & Build 360° Intelligence Profile Forestry Domain

Project Overview We are building Fore_STA, a B2B intelligence platform for the Romanian forestry and wood processing market. We do not need a technician who waits to be told what chart to make. We need a Lead Analytics Architect to design the end-to-end commercial business logic, architect the PostgreSQL dimensional layers, and implement a Unified Company 360° Profile in Apache Superset. The objective is to reveal a wood-processing company's complete operational reality on one screen: what timber they bid on, what they actually win and at what prices, what physical timber they transport out of forests (via SUMAL 2.0 / APVs), and whether their balance sheet supports their market activity. Key Responsibilities & Deliverables 1. Business Logic & KPI Formulation o Define the composite indicators that expose market behavior:  Bidding Aggression Index: Average spread paid over starting price (pret_adjudecare - pret_pornire) by assortment and species.  Procurement Footprint: Sourcing concentration across forest districts (Romsilva vs. private ocoale).  Supply Chain Audit Ratio: Discrepancy metric comparing won auction volumes ($m^3$) vs. transported timber volumes (SUMAL $m^3$) vs. reported financial turnover.  Financial Capacity Tier: 10-year revenue/debt trends evaluated against raw material purchasing spend. 2. Data Modeling & Stacking (PostgreSQL Layer) o Build clean, performant SQL views and materialized views that unify:  Corporate records (companies_company: CUI, financials, legal status).  Auction records (auctions_lotadjudication, auctions_timberlot).  Physical forestry records (APVs, UAT locations, transport permits). o Implement entity matching logic (reconciling company names to CUIs and adjudications to announcements via deterministic and fuzzy parameters). 3. Superset Dashboard Experience (The Company 360° Snapshot) o Architect an intuitive dashboard layout:  Top Banner: Company card, fiscal status, 10-year financial trendlines.  Procurement Dynamics: Win-rate gauges, price-per-species heatmaps, auction frequency.  Physical Sourcing Basin: Geospatial deck.gl map showing where timber originates (UAT / Ocol).  Sanity-Check Audit: Dual-axis visual comparing official revenue against market wood acquisitions. o Configure global cross-filters (Company search, Year, Species, Region) with sub-3-second query execution. 4. Documentation & Maintenance Protocol o Provide a concise architecture map explaining how views are stacked and how materialized views refresh. Candidate Requirements • Analytics Engineering Mindset: Proven ability to define metrics and business logic from scratch, not just execute specs. • Mastery of Apache Superset: Expert in dataset semantic setup, calculated expressions, layout design, and query optimization. • Advanced PostgreSQL: Materialized views, window functions, CTEs, JSONB manipulation, indexing (GIN/BTREE), and fuzzy matching (pg_trgm). • Strong Advantage: Understanding of Romanian forestry data (SUMAL 2.0, APVs, Romsilva auctions, Ocoale Silvice) or native/professional Romanian language skills (raw records and PDF tables are in Romanian).