A detail-driven Data Strategist and Systems Engineer with a rigorous foundation in corporate finance. I specialize in translating complex operational datasets into actionable intelligence, building high-performance local data engineering utilities, and securing software workflows.

• 📊 What I Do: Data Analyst & Analytics Specialist specializing in Python, SQL, Power BI, and local-first data engineering tools.
• ⚡ Currently Building: Automated data pipelines, issue-tracking scrapers, and SQL performance benchmarks (data-engine-benchmarks, duck-diff).
• 🐙 Open Source: Active contributor across major data engines, AI/RAG architectures, and mathematical libraries (langchain, sympy, sqlfluff, ibis-project, scikit-learn).
• 🌱 Currently Exploring: Advanced DuckDB optimizations and high-performance local data tooling.
• 🤝 Open to: Roles in Data Analytics, Analytics Engineering, or Open-Source collaborations.

I actively contribute to the core infrastructure of Python data frameworks, AI/RAG architectures, symbolic mathematics, and compiler engines:

🤖 LangChain | Vector Store Integrity & Schema Ingestion — View PR #40079
Context: Ingestion pipelines allowed mismatched lengths between input text documents and provided ids, leading to silent indexing corruption in downstream vector stores (Issue #36203).
What I Built: Implemented strict assertion and validation checks in VectorStore.add_texts ensuring equal dimensions between inputs and ID sequences prior to embedding dispatch. Added unit tests verifying defensive error raises on dimension mismatches. (Stack: Python, LangChain, Vector Embeddings, Pytest)
Direct Proof: PR #40079

🧠 Semantica | FAISS Persistence & Markdown Parsing — View PR #1314 • PR #1315
Context: Serializing FAISS indexes dropped explicit vector_ids and custom metadata mappings during save/load operations (Issue #1272). Concurrently, the markdown normalizer was stripping intentional whitespace within fenced code blocks.
What I Built: Engineered persistent metadata storage alongside FAISS indexes to guarantee 1:1 ID parity across I/O cycles. Patched document normalizers to preserve indentation within code fences. Authored contributor guidelines to streamline external contributions. (Stack: Python, FAISS, Vector Storage, AST Normalization)
Direct Proof: PR #1314 • PR #1313 • PR #1315 • PR #1386

🔍 SQLFluff | AST Parser Engine & Dialect Parity — View PR #8396
Context: DuckDB dialect queries using SUMMARIZE encountered grammar classification mismatches against the unreserved keyword tree.
What I Built: Corrected keyword classifications in dialect_duckdb.py by relocating SUMMARIZE to reserved_keywords. Updated dialect fixtures, regenerated the fixture parse tree YAML, and ensured full compliance across all 9 pre-commit static analysis checks (mypy, ruff, yamllint, doc8, codespell). (Stack: Python, SQLFluff AST, DuckDB, Pre-commit)
Direct Proof: PR #8396 • Commit 0397f6b

📊 Ibis | Cross-Engine Type Translation & Schema Inference — View PR #12086
Context: Ingesting Polars tables with mixed or custom object types triggered unhandled KeyError: Object exceptions during schema reflection.
What I Built: Extended PolarType.to_ibis to map both class and instance forms of pl.Object to dt.Unknown(nullable=True). Designed a standalone client integration test in test_client.py using direct ibis.polars.connect() memtables to safeguard schema inference without external dataset dependencies. (Stack: Python, Polars, Ibis, Pytest)
Direct Proof: PR #12086

🛡️ py-simple-wrap | SQL Injection Mitigation & CRUD Testing — View PR #199
Context: Database helper methods lacked validation guards against malicious input payloads in table and column identifiers.
What I Built: Hardened database helpers with regex whitelisting (_check_if_valid) and mandatory parameterized tuple execution to block stacked queries and UNION-based injection attacks. Synced updates with upstream additions (run_update, ExperimentalWarning), restored module re-exports in __init__.py, and authored a comprehensive 52-test pytest harness. (Stack: Python, SQLite3, Pytest, Application Security)
Direct Proof: PR #199

📚 icalendar | API Documentation Architecture & Ruff Linting — View PR #1749
Context: Public APIs across core calendar parsing and serialization engines contained inconsistent docstring formatting that violated community style specifications.
What I Built: Refactored docstring headers to adhere to formal Sphinx/PEP Parameters: conventions across parser_tools.py and cal/component.py (resolving issue #1072). Enforced code cleanups across automated Ruff formatting and linting pipelines with zero regressions. (Stack: Python, Ruff, Sphinx Docs, RFC 5545)
Direct Proof: PR #1749

🧠 Vector Ingestion & Storage Architecture

• LangChain (PR #40079) — Enforced strict dimensional boundary assertions in VectorStore.add_texts, eliminating silent metadata drift and corrupt indexing across text chunk ingestion pipelines (closes #36203).
• Semantica (PR #1314 • PR #1313) — Resolved serialization state desynchronization in FAISS vector stores by establishing atomic vector_id and document metadata persistence routines across index read/write cycles (closes #1272).

🔍 AST Parser Engines & Dialect Parity

• SQLFluff (PR #8396) — Diagnosed and reconciled DuckDB SQL grammar mismatches by reclassifying SUMMARIZE into dialect-reserved keywords, regenerating AST syntax tree YAML definitions, and passing an automated 9-hook static verification pipeline (mypy, ruff, yamllint, doc8, codespell).

📊 Cross-Engine Type Translation & Schema Inference

• Ibis (PR #12086) — Eliminated unhandled KeyError: Object exceptions by engineering schema reflection logic that maps Polars pl.Object structures cleanly to dt.Unknown(nullable=True) with standalone memtable integration tests.
• SymPy (PR #50354) — Modernized type hints across sympy.utilities.decorator to reinforce strict static type checking and type inference during symbolic evaluations (closes #28806).

🛡️ Application Security & SQL Injection Mitigation

• py-simple-wrap (PR #199) — Hardened database utilities against stacked queries and UNION-based injection attacks via regex identifier whitelisting (_check_if_valid) and mandatory parameterized tuple execution; authored a 52-test pytest harness.
• Adexa (PR #10) — Designed 300+ lines of targeted pytest suites validating automated AI remediation engines against common web vulnerability vectors.

⚡ Protocol Standardization & Pipeline Hardening

• MCP Migrate (PR #254) — Hardened RD10 wire boundaries for Model Context Protocol migrations and introduced the ServerDiscoverFixer test harness.
• HFlow (PR #281) — Refactored data curation pipelines to isolate single-SELECT validation rules directly within core libraries (closes #280).
• Scitex IO (PR #166) — Built round-trip binary Feather serialization validation checks and bad-extension error handlers.

🔀 Advanced Version Control & Upstream Reconciliation

• Git Workflows — Maintained high branch hygiene across high-velocity multi-contributor upstream repositories using interactive rebasing, multi-file 3-way merge conflict resolution, hook automation, and safe force pushes via --force-with-lease.

High-performance, air-gapped utilities built for low-latency developer workflows, local data reconciliation, and automated codebase governance:

⚡ Tool	🎯 Focus & Architecture	🛠️ Tech Stack	🔗 Status
duck-diff	Embedded table & schema reconciliation engine for fast multi-million-row dataset diffing without cloud transit.	DuckDB Python Arrow
data-engine-benchmarks	Automated benchmarking matrix measuring throughput, peak memory usage, and vector speed across data engines.	Polars DuckDB Pandas
repo-doctor	Local-first codebase scanner auditing leaked secrets, software licenses, dependency debt, and AI context hygiene.	Python AST Git Hooks
sqlean-lint	Lightweight static analysis engine detecting SQL anti-patterns, non-SARGable predicates, and join cartesian risks.	Python SQLFluff CLI
dbt-optimizer	Dependency DAG and compilation cost analyzer highlighting model bloat and high-latency transform paths.	dbt-core Python NetworkX
github-issue-hunter	High-speed aggregator and telemetry board indexing and filtering open issues across high-volume open-source repos.	Python REST APIs SQLite

• 🏫 Power BI Developer & Instructor | PNY Trainings (NAVTTC) (Feb 2026 – May 2026)
  • Engineered and delivered comprehensive technical curriculum in Power BI, SQL, and Python; established best practices for ETL pipelines and advanced data modeling.
• 🍕 Junior Data Scientist | Timmy's Pizza (Nov 2023 – Dec 2025)
  • Optimized local delivery routes, staffing schedules, and supply chain visibility using Pandas and interactive Power BI dashboards.
• 💻 Data Analyst Intern | PNY Trainings (June 2025 – Present)
  • Spearheaded a data-driven marketing analysis for a key e-commerce client, leveraging Power BI and Advanced Excel to project a 15% increase in customer engagement.
  • Automated a reporting pipeline using Python (Pandas) and SQL, reducing manual data processing for weekly sales reports by 10 hours per month and improving efficiency by 30%.
• ⛽ Procurement Intern | Sui Northern Gas Pipelines Limited (SNGPL)
  • Analyzed vendor performance metrics for a portfolio of 50+ suppliers, creating KPI dashboards in Excel that contributed to an estimated 5% reduction in procurement costs.
  • Streamlined the digital record-keeping process for Purchase Orders, designing a new workflow that reduced document retrieval times by over 50%.

• 🎓 Bachelor of Science in Accounting and Finance — Hailey College of Commerce, University of the Punjab
• 🏆 IBM Data Analyst Professional Certificate
• 🏆 Google Advanced Data Analytics Professional Certificate
• 🏆 IBM Data Science Professional Certificate
• 🏆 IBM Data Engineering Professional Certificate
• 🏆 Google Digital Marketing & E-commerce Professional Certificate

[ Git Push / PR ] ──► [ GitHub Actions ] ──► [ Linting & PyTest Matrix ] ──► [ Deploy / Publish ]