Data Integration Specialist

Gumi | Intelligent Ventures

Important: Must be based in Kenya

Responsibilities:

• You'll take messy, fragmented enterprise data and turn it into a structured foundation that AI systems can work with. This means connecting client systems (SharePoint, CRMs, fund administration platforms, email, spreadsheets), ingesting and processing documents at scale, building data pipelines, mapping permissions, and ensuring data quality across everything the AI layer touches.
• You'll work directly with client data teams to understand their systems, access their data, and build the integration layer that everything else depends on. If the data foundation isn't right, nothing downstream works. You own that foundation.

Requirements:

• 3–6 years experience in data engineering, backend engineering, or systems integration
• Strong Python skills for data processing (pandas, document parsing libraries, scripting)
• Experience connecting enterprise systems via APIs, database connectors, and file-based integrations
• Comfort with messy, real-world data — PDFs that don't parse, spreadsheets with merged cells, inconsistent formats across sources
• Familiarity with vector databases and embedding pipelines (Pinecone, Weaviate, Qdrant, or similar) — or willingness to learn quickly
• Understanding of data access controls and permission mapping
• Ability to work independently with minimal supervision once pointed in the right direction
• Experience with financial services data — fund administration systems, portfolio management platforms, banking or fintech infrastructure
• Experience working in African enterprise environments where the tech stack is a mix of enterprise software, local solutions, and spreadsheets
• Previous work parsing and structuring large document libraries (contracts, reports, memos)

Engagement structure: Contract, project-by-project. Typical engagement: 8–12 weeks. Heaviest workload in weeks 1–5. Remote-friendly with occasional client site visits (1–3 days per engagement). Must be available during GMT+1 to GMT+3 working hours.