ABP new specialization
Data Lakehouse and RAG systems for real business needs
We combine experience in process automation with modern data practices: from storage and analytics to intelligent search and generation.
Core directions
- Data Lakehouse: analytics, data marts, Time Travel
- RAG platforms: search + generation on corporate data
- Legacy: HCL Notes/Domino support
About the company
ABP (BPMDoc) is an engineering team that has been helping businesses automate processes and develop corporate systems since 2005. Our primary focus used to be the Domino platform; today we concentrate on data and analytics.
What we do
- Ready-made document workflow automation solutions
- Custom software to improve process efficiency
- Support and evolution of HCL Notes/Domino solutions
2026 focus: data and intelligence
Data Lakehouse
We combine the advantages of DWH and Data Lake: transactions, scale, and fast changes.
RAG systems
Intelligent search across corporate knowledge and answers based on your data.
Domino legacy
We maintain and modernize critical solutions on HCL Notes/Domino.
Why Data Lakehouse
Challenges of classic DWH
- Expensive and slow ETL/model changes
- Data stays in source systems or gets lost
- No precise view “as of a past date”
- Hard to prepare raw data for ML and Data Science
Lakehouse solves this
- ACID and Time Travel on top of Data Lake
- Schema-on-read flexibility without quality loss
- Fast changes and ad-hoc analytics
- Unified data for BI and ML
Data Warehouse
High performance, but costly changes and low flexibility.
Data Lake
Flexibility and raw data for ML, but no transactions and slow SQL.
Lakehouse
Transactions, Time Travel, one platform, fast time-to-analytics.
Data marts
For heavy BI workloads we build data marts on high-performance engines synchronized with primary sources.
Query optimization
Tables are created only for real needs and required volume.
High concurrency
ClickHouse and MPP engines sustain hundreds of concurrent queries.
Synchronization
Marts are refreshed regularly from primary data.
RAG systems
We build Retrieval-Augmented Generation solutions that combine search across corporate data with response generation. This helps teams get accurate knowledge without manual search.
Typical pipeline
- Data ingestion and knowledge cleansing
- Vectorization and indexing
- Prompt orchestration and quality control
- Integration into portals, bots, and BI
Lakehouse technology stack
MinIO / S3
Reliable object storage with access control and scalability.
Apache Iceberg
Transactions, partitioning, and Time Travel on Data Lake.
Trino / Spark
SQL analytics over Lakehouse data and source integration.
ClickHouse
High-performance marts and concurrent BI workloads.
Apache NiFi
ETL and data ingestion with delivery guarantees.
dbt
Managed SQL transformations and model documentation.
How we work
Discovery
Assess current sources, target metrics, and constraints.
Pilot
Build a PoC fast and confirm value on client data.
Implementation
Deploy the platform, set up marts, and automate workflows.
Support
Maintain, evolve, and train client teams.