Data Engineer | Technical Business Analyst | Capital Markets & Regulatory Reporting
📍 Bengaluru, India | 📧 sujeetsinghjsr@gmail.com | 🔗 LinkedIn | 💼 Publicis Sapient
I design and deliver enterprise-scale data pipelines for regulated industries — capital markets, energy, and financial services.
- 🏗️ Currently building a MiFID II Regulatory Data Pipeline at Publicis Sapient
- 🌐 Previously led Shell's Data Mesh transformation at IBM
- 📊 14+ years across MiFID II · CFTC · BCBS 239 · G20 regulatory frameworks
- ⚙️ Specialist in pipeline architecture · trade linkage · delta check engines · data persistence
Publicis Sapient | Jan 2026 – Present
Building a production Regulatory Data Pipeline (RDP) for MiFID II ARM & APA reporting. Translating Murex MX3 trade events into regulatory submissions via Kafka and AWS.
Murex MX3 → De-Dup → MX Message Type Identifier → Trade Event Enricher
→ GT Enrichment (LEI · MIC · IDM · EDM)
→ JSON Data Products Creator (35+ MxML → 1 flat JSON)
→ JSON FX SWAP Aggregator → Post-Filter → ANNA Data Enricher
→ [TRADE LINKAGE] → [DATA PERSISTENCE]
→ Jurisdiction Eligibility → Outbound Reporting
→ Aggregator / Delta Check (Phase A → B → C → D)
→ ARM (batch) + APA (real-time) → Trade Repository
- Designed NB / CREATOR_NB waterfall lookup replacing CONTRACT_ID approach
- Built replay detection using TRADE_REFERENCE + TMIT uniqueness check
- Designed DSL data extraction layer — XPaths in Excel on AWS S3, hot-reload on restart
- Handled PF / NPF / NRPT MiFID event categories and IS_NRPT propagation
- Produced: Jira ticket (AC, DoD, XPath details), draw.io flow, DSL rules file, UAT pack
- Table:
RDP_TRADE_DATA_LINKAGE— TRADE_REF · CREATOR_TRADE_ID · MIFID_LINK_ID
- Designed two-table architecture from Kear Chea (design authority) clarification call
- Table 1:
RDP_MIF_ELIGIBILITY_CACHE— regime-agnostic · IDM + EDM only · cache layer - Table 2:
RDP_MIFID_TRADE_DATA_PERSISTENCE— MiFID-eligible trades only - 8 DSL rules: SOFT_LINK (MIC F36, Venue ID F3) · CARRY_OVER (TRN F2, Trading Date F28, IDM F57, EDM F59) · DELTA_DRIVEN via Phase C write-back (Qty F30, DECR_INCR F32)
- Phase A: polls persistence WHERE STATUS = PENDING
- Phase B: jurisdiction eligibility gate (IS_NRPT = Y → SUPPRESSED)
- Phase C: delta check on outbound table → write-back Qty + DECR_INCR + MIFID_ACTION
- Phase D: ARM/APA submission → MIFID_STATUS = COMPLETE
- MIFID_ACTION: NULL → NEWT / REPL / CANC
- Dimensional coverage approach — 43 financial products × 21 event types
- Product × Event matrix eliminates combinatorial explosion
- P0 scenarios cover NEWT/REPL/CANC, NPF carry-forward, NRPT propagation, replay detection
Tech stack: Murex MX3 · Kafka · AWS S3 · DSL Framework · JSON · SQL · ARM/APA · MiFID II RTS 22
IBM India embedded at Shell | Aug 2022 – Oct 2025
Led Shell's enterprise Data Mesh transformation across global business units.
- Designed Databricks Lakehouse using Delta Lake · Parquet · ADLS Gen2
- Implemented Unity Catalog + Collibra for data governance · lineage · metadata
- Built data products on Shell.ai platform for engineers · analysts · data scientists
- Delivered GDPR-compliant data modernisation using CDC · batch · real-time streaming
- Led PoC initiatives validating capabilities before enterprise rollout
Tech stack: Databricks · PySpark · Apache Spark · Azure · AWS · Delta Lake · Parquet · ADLS Gen2 · Unity Catalog · Collibra · Data Mesh · Data Products · CDC
JPMorgan Chase & Co | Oct 2016 – Jul 2018
- Built ingestion and standardisation pipeline for 4M+ financial instruments
- Bloomberg Data License + Python scripts → Global Instrument Master (GIM)
- 99% data integrity for client golden source systems
- Handled corporate actions: splits · dividends · listings · delistings · ticker changes
Tech stack: Python · SQL · Bloomberg Data License · ISIN · LEI · Reference Data
Standard Chartered Bank (2018–2022) — Lead BA, Financial Markets Middle Office
- Led Murex migration (FEDS → Murex) — $3M cost savings · 35% processing improvement · 95% error reduction
- FX Spot · Forward · Swaps · NDF trade booking to risk systems
S&P Global (2012–2016) — BA, Real-time Market Data
- Market data direct feed products · daily quality alerts · corporate action reconciliation
| Category | Technologies |
|---|---|
| Data Engineering | Databricks · PySpark · Apache Spark · Delta Lake · Parquet · CDC |
| Cloud | AWS (S3 · Cloud Practitioner certified) · Microsoft Azure (AZ-900 certified) |
| Streaming | Kafka · Event Mesh · Real-time Streaming · Batch Processing |
| Storage / Format | Delta Lake · ADLS Gen2 · JSON · MxML · XML · FpML · FIX Protocol |
| Governance | Unity Catalog · Collibra · Data Lineage · Metadata · Data Catalog |
| Regulatory | MiFID II ARM/APA RTS 22 · CFTC LTR · BCBS 239 · MiFIR · GDPR |
| Domain Systems | Murex MX3 · Bloomberg RHUB · Trax · OTCR · ANNA DSB · RDH/RDP |
| Database | SQL · Schema Design · DDL · Indexing · Data Quality · Reconciliation |
| Frameworks | Data Mesh · Data Products · Self-Serve Platforms · Agile Scrum |
| Tools | JIRA · Confluence · draw.io · Python · Excel (Dynamic Matrices) |
| Repository | Description |
|---|---|
| mifid-rdp-pipeline-design | MiFID II RDP trade linkage and data persistence architecture |
| data-mesh-shell | Shell enterprise Data Mesh platform design patterns |
| delta-check-engine | Delta check state machine for MiFID II regulatory reporting |
| regulatory-uat-framework | 89-scenario UAT test pack using dimensional coverage |
| reference-data-pipeline | JPMorgan 4M+ instrument reference data pipeline |
| databricks-lakehouse-patterns | Databricks Delta Lake patterns from Shell.ai platform |
- ☁️ AWS Cloud Practitioner — Amazon Web Services
- ☁️ Microsoft Azure Fundamentals (AZ-900)
Open to Data Engineering roles — pipeline architecture · regulatory data · cloud data platforms
📧 sujeetsinghjsr@gmail.com | 🔗 LinkedIn