Enterprise data platforms, document AI, GenAI copilots and applied industry AI — engineered on AWS, grounded in the SAP data we've been running for three decades.
Enterprises don't need another vendor promising GenAI. They need someone who can land the data foundation, get documents into structured records, ground copilots against source-of-truth, and ship the applied products that show a return.
That's what Phoenix runs. AWS as the platform. SAP as the data heritage. A track record of moving customers from PoC decks to production — including a Smart OCR product live in AWS Marketplace and a pricing-intelligence product deployed in real estate.
AWS-first data and AI built on three decades of SAP-data context. The frontier we invest in — with proof already live: a lakehouse practice with 70%+ archiving lifts, Smart OCR in AWS Marketplace, an AI real-estate product in production.
A lakehouse on AWS with SAP Datasphere and BW as governed sources, engineered by the team that already knows your SAP data.
ProvenArabic and English extraction from IDs, invoices, contracts and handwriting — structured JSON in S3, feeding Bedrock knowledge bases.
Deployed · AWS MarketplaceRAG knowledge bases and enterprise copilots on Amazon Bedrock; SAP Joule enablement where the data lives in SAP.
Frontier · building proofIndustry-tuned AI products — pricing intelligence (EstateEdge, retail price-compare) and logistics AI for routing, fleet and driver performance.
Product liveA cross-section of what Phoenix has shipped under the Data & AI practice. Each one lives on the customer's own AWS estate, integrated with the systems that already run the business.
Tender documents → ranked opportunities in minutes.
Conversational AI reads incoming tender packs, extracts requirements, flags risks and produces evaluation memos. Sales and pre-sales teams read the memo, not the RFP.
See the GenAI & Agents track →Onboarding, recruitment and policy queries under one agent.
Employee-facing copilot grounded in HR policy and integrated with the HRIS — handles onboarding checklists, benefits questions, recruitment triage and performance workflows.
See the GenAI & Agents track →Competitor moves and market signals in real time.
Streaming intake of competitor listings, campaigns and market signals; automated content drafts and campaign insight for marketing teams working against fast-moving verticals.
See the Applied AI track →Raw operational data → narrative strategic reports.
Scheduled generation of the kind of report an analyst spends a week producing — turned into a minutes-long automated flow with citations back to source data.
See the GenAI & Agents track →Concept-to-production cycle shortened for a design house.
Industry-tuned generative AI over the customer's own design corpus — reference-augmented ideation with production-ready outputs. An applied-AI build for a specific vertical.
See the Applied AI track →Model Context Protocol — CRM ↔ frontier-model translator.
Enterprise MCP layer connecting the customer's CRM (and adjacent systems) to frontier-model surfaces. Copilots read authoritative live data instead of stale exports.
See the GenAI & Agents track →Arabic and English extraction from IDs, invoices, contracts and handwriting. Structured JSON in S3, feeding Bedrock knowledge bases. Sold as a product, deployed for customers on their own AWS accounts.
See the product →AWS-native real-estate pricing product. Sourced from live listings, normalised, benchmarked, and served as insight to developers and brokers across the region.
See the product →Long-cycle migration and archiving programmes on SAP estates. Historic data moved to Amazon S3 tiered storage, hot data kept in HANA. Same-application UX, lower run cost.
AWS is the platform of record — Bedrock for GenAI, S3/Redshift for the lakehouse, Marketplace for productized solutions. Where the workload sits in SAP, SAP Joule is enabled directly. Non-AWS clouds are supported case-by-case where a customer standard requires it.
Most data foundations fail on messy source integration. Three decades of running SAP end-to-end means Phoenix reads S/4HANA, BW and Datasphere as first-class sources — extractions arrive clean, semantics are preserved, and downstream ML models train against data the business actually trusts.
Smart OCR — a document-AI product live on AWS Marketplace, extracting Arabic and English into structured JSON. EstateEdge — an AWS-native real-estate pricing product used by developers and brokers. SAP archiving programmes that have cut storage footprints by 70%+.
Yes — that's the common pattern. Smart OCR ingests the corpus into structured JSON in S3, Bedrock Knowledge Bases wrap it with retrieval, and a copilot layer sits on top with guardrails, citations and access control. Grounded, not hallucinated.
Each applied-AI product is deployed as a tuned instance per customer — same core engine, customer-specific data ingestion, KPIs and interfaces. Sold via subscription against a named business KPI, not licence fees.
Whether you're scoping an SAP move to cloud, restarting a stalled programme, or just trying to figure out where data and AI fit — start with a conversation.