Sabbir HossainData Engineer
Career Ledger

Career Ledger

Bell is where I'm focused now. Hopkins and UofT built the depth.

Ledger mapCareer lineage & dependencies
Drawn byS. Hossain
ScaleNTS
RevisionB
RefSH-CL-002
Lineage (time)
2026+
2025
2024
2023
2022
2021
2020
2019
2018
The pivot → CS + BioInfo
ABRCMS '23 · oral award
Harvard NCRC · plenary talk
BSc Hons · 3.96 major GPA
ABRCMS '24 · poster award
Technical Gatekeeper · 3 mo
3 lead-author manuscripts
Fig. 02.1Career lineage graph
4 nodes · 4 dependencies · 2018 → presentCareer Ledger · Quiet Systems · Rev B

Bell Canada · Data Engineer

June 2025 - Present

I own production data systems that millions depend on every day.

80,000discrepancies reconciled
800%coverage expansion
~3.5Mduplicate rows eliminated
  1. 01DetectAnomaly detected in record counts and query latency.
  2. 02TraceTraced lineage across pipelines, logs, and upstream sources.
  3. 03RepairRepaired transformations and backfilled missing partitions.
  4. 04ReconcileValidated counts and reconciled datasets end-to-end.
  5. 05MonitorAdded data quality checks and latency monitoring.

the validator saw it before the dashboards did. ~28,000 mismatches.

Ownership signals
  • Enforce exact row-count parity before every downstream step.
  • Reverse-engineer undocumented legacy logic instead of stalling on poor docs.
  • Rebuild fragile reporting views into maintainable, validated ETL.
  • Restore data trust by recovering lost history and validating the backfill.
  • Treat throughput, correctness, and observability as one reliability problem.
  • Separate technical feasibility from access and process blockers.
  • Prove a warehouse migration row by row before calling the copy correct.
Runbook excerpt (generalized)
  1. 01# Incident: rows multiplying on load
  2. 02# Runbook: dedup_recovery (generalized)
  3. 03 
  4. 041. Confirm the duplication
  5. 05 - Count distinct vs total per key
  6. 06 - Sample the shape of the dupes
  7. 072. Find where it enters
  8. 08 - Trace lineage to the load step
  9. 09 - Diagnose the load-key mismatch
  10. 103. Re-key the load
  11. 11 - Upsert key vs primary index
  12. 12 - Make the write idempotent
  13. 134. Prove it before promoting
  14. 14 - Idempotency: run it twice
  15. 15 - Reconcile counts end to end
  16. 165. Stage the cutover
  17. 17 - Rehearse a reversible plan
  18. 18 - Monitor after go-live
Query plan (abridged)
Before
Rows
23,000,000
Runtime
12 min
Reduction
baseline
  • Dynamic runtime computation
  • Sequential event sequencing
  • Hardcoded 30-day lookback
  • Non-idempotent retries
  • Monolithic SQL
After
Rows
23,000,000
Runtime
2 min
Reduction
6×
  • Materialized pre-aggregation
  • Two-pass group-by propagation
  • SCD Type 2 temporal joins
  • Composite upsert idempotency
  • Zero-downtime abstraction layer

Same 23M-row result set. 12 min → 2 min.

Ownership matrix (current role)
AreaI ownI partnerI support
Data Pipelines✓
Data Quality✓
Incident Response✓
Backfills & Recovery✓
Performance Tuning✓✓
Stakeholder Comms✓✓
Monitoring & Alerting✓✓
Documentation✓✓✓
Open System Note 01 →
Verified signals
Enterprise analytics platform delivery
MongoDB → BigQuery migration delivery
Methods (compact)
  1. IngestCollect raw data at scale
  2. ValidateEnforce quality & provenance
  3. ModelBuild models that explain & generalize
  4. ReproduceRe-run, stress, and verify
Data flow (abridged)
  1. Ingest (Raw)
  2. Validate (QC + Schema)
  3. Model (Analysis)
  4. Reproduce (Re-run)
  5. Outputs (Artifacts)

Schema ref: JHPM-CORE-01 · Versioned · Traceable · Reproducible

Awards

Outside proof that the work held up in front of real review.

yearrecognitionissued byref
2024Poster Presentation AwardReceived top honors for a graduate-level poster presentation among 150+ graduate presenters on advancing open-source bioinformatics platforms.Annual Biomedical Research Conference for Minoritized Scientists (ABRCMS) - Graduate Division$2,500 travel awardproof ↗
2024Friends of Arts and Science AwardAwarded for academic excellence in Computer Sciences and Physical & Life Sciences during the 2023-2024 academic year.University of Toronto - Faculty of Arts & Science
2024Plenary SpeakerSelected as 1 of only 12 plenary speakers from over 5,000 national applicants. Delivered a keynote on integrating transcriptomics and proteomics data for glioblastoma research.National Collegiate Research Conference (NCRC) - Harvard Universityproof ↗
2024Poster PresentationPresented computational approaches for cancer biomarker identification using integrated multi-omics analysis and machine learning methodologies.National Collegiate Research Conference (NCRC) - Harvard Universityproof ↗
2023Oral Presentation AwardAwarded top presenter in the Computational and Systems Biology division, selected from 80 oral presenters at a conference with over 3,500 attendees.Annual Biomedical Research Conference for Minoritized Scientists (ABRCMS) - Computational and Systems Biology Division$2,500 travel awardproof ↗
2023Friends of Arts and Science AwardAwarded for academic excellence in Computer Sciences and Physical & Life Sciences during the 2022-2023 academic year.University of Toronto - Faculty of Arts & Science
2022Friends of Arts and Science AwardAwarded for academic excellence in Computer Sciences and Physical & Life Sciences during the 2021-2022 academic year.University of Toronto - Faculty of Arts & Science
Education

Strong CS and bioinformatics foundations, plus a lot of range.

University of Toronto

St. George Campus · Toronto, Ontario, Canada
Degree
Bachelor of Science (Honours)
Program
Computer Science, Bioinformatics & Computational Biology
Minor
Immunology
Major GPA
3.96 / 4.0
Graduated
June 2024
Sabbir Hossain · Career LedgerPage 02