Mactores is the agent-native AWS modernization firm. Most modernization work doesn't ship, it stalls in pilots, slips a year, or lands at three times the budget. We exist to ship it: production systems running, legacy retired, outcomes measured. Our delivery is built on Aedeon, the agent platform built by Mactores' founders' sister company, which absorbs the repetitive 60–70% of engagement work, discovery, dependency mapping, validation, test generation, that traditional consulting bills human hours against. Forward-deployed engineers own the rest: architecture, judgment, and cutover, on dates we commit to in the contract.

Enterprise SAP data is where data platform modernization goes to stall. The tables are cryptic, the semantics live in someone's head, and the extraction tooling punishes wrong choices months later. This role exists because getting that data out cleanly and back in usably, takes an engineer who knows SAP from the inside.

As our SAP Data Engineer, you'll own the SAP side of a modern data pipeline for enterprise customers running SAP on AWS. You'll design and ship the extraction layer that lifts data out of SAP HANA into S3 cleanly and incrementally, and you'll build the BW / BW4HANA load path that takes curated data back into the warehouse for business users to query. Agents absorb the repetitive discovery and validation work around you; the judgment calls which extraction tool per source, how the semantics map, what the S3 contract looks like, are yours.

You'll work alongside a PySpark engineer who owns the transformation layer on EMR, meeting them at the S3 boundary. You'll be the SAP voice on customer calls, and the person engineering counts on when raw SAP tables need to become datasets a business analyst can trust. This is a freelance engagement, remote from India, embedded with a customer team through cutover.

What you'll do

• Build extraction pipelines from SAP HANA to AWS S3 using SLT, ODP, CDS views, SDI, and native HANA SQLScript — picking the right tool per source and per latency requirement.
• Model raw SAP tables across FI/CO, MM, SD, and adjacent modules into clean, semantically meaningful datasets the downstream Spark layer and business users can actually use.
• Design and operate delta and CDC patterns so incremental loads stay correct, idempotent, and replayable.
• Write ABAP extractors where standard SAP tooling falls short, and document them so future engineers can change them safely.
• Own the write-back path: load curated data from S3 into SAP BW / BW4HANA and model it for end-user reporting and analytical querying.
• Land data in S3 as Parquet with sane partitioning, schemas, and IAM scoping, and define the contract with the PySpark engineer at the ingestion-to-transformation boundary.
• Embed with a customer team, ship the pipeline to production, and stay close enough through cutover to know it actually runs.

What we're looking for

• Deep SAP HANA extraction experience: real production work with SLT, ODP / operational data provisioning, CDS views, SDI, and HANA SQLScript.
• A strong grasp of SAP table structures and the business semantics behind them in at least one functional area (FI/CO, MM, SD, or similar) you can turn raw tables into models a business analyst recognizes.
• Solid delta and CDC instincts: you've designed incremental loads that survive reprocessing, late-arriving data, and source-side schema drift.
• ABAP fluency sufficient to build custom extractors when standard tooling can't reach the data.
• SAP BW / BW4HANA data loading and modeling experience for the consumption side of the pipeline.
• Working AWS knowledge: S3 landing zones, Parquet, basic IAM enough to collaborate confidently on an AWS-hosted pipeline without needing to own the platform.
• Strong written and spoken English. You'll be on customer calls and working across geographies.

You'll be preferred if you

Have worked with AWS Glue Data Catalog or similar metadata layers over S3.
Have shipped SAP-to-cloud data lake patterns before, with opinions on what broke and how you'd do it differently.
Have done client-facing consulting or forward-deployed delivery, not just internal IT work.
Are comfortable collaborating with a Spark-based transformation layer at the S3 boundary, even if you don't write the Spark yourself.

How we work?

We embed with the customer's team and own outcomes through production not through the slide deck. Agents handle the repetitive majority: discovery, mapping, validation, test generation. Engineers own what agents can't: architecture, semantics, risk tradeoffs, cutover. We commit to delivery dates and absorb overage cost for delays inside our control. That model only works when the person who designed the pipeline is still there when it cuts over which is why this role stays close through go-live.