Adyan Ahmad Quazi

Adyan Ahmad QuaziData and strategy analystGreater Toronto Area

I measure what the data can support, then show the working.

About three and a half years in analyst roles: Excel and Power BI reporting for an executive team, then the market-entry research program at a carbon-fibre composites and additive-manufacturing startup. I have published six analyses in SQL and Python on public data, and a CI job that re-runs every figure five of them quote, 919 in all, against a fresh pull of each source.

See the work Portfolio repository adyan.ahmad@gmail.com

What repowering did to US wind plants

Repowered plants against never-repowered plants of the same vintageLine chart, years since repowering from minus 4 to plus 5 on the horizontal axis, generation relative to the year before repowering in percent on the vertical axis. Before the work the estimates are +3.9, +2.3 and minus 4.5 against a base of 0; the repowering year is minus 6.7; then +33.9, +46.6, +54.7, +59.4 and +61.6 from year 1 to year 5. A shaded 95% interval is drawn from the repowering year onward, the only years for which the memo prints bounds. A dashed line marks the naive regression at +30.4. Headline: +48.4% from the first full year, interval +36.3% to +61.4%.Generation relative to the year before repowering, %−20020406080−4−3−2−10+1+2+3+4+5Years since repoweringcohortsplants780780780780780780673567461338naive regression: +30.4%-4: +3.9-3: +2.3-2: −4.5-1: base year, 0 by construction0: −6.7, 95% interval −14.2 to +1.5+1: +33.9, 95% interval +20.9 to +48.1+2: +46.6, 95% interval +34.4 to +59.8+3: +54.7, 95% interval +40.2 to +70.6+4: +59.4, 95% interval +43.5 to +77.0+5: +61.6, 95% interval +41.6 to +84.4before the work: placebo p = 0.268repowering year −6.7+48.4% from the first full year after the work95% interval +36.3% to +61.4%
Repowered plants against never-repowered plants of the same vintageLine chart, years since repowering from minus 4 to plus 5 on the horizontal axis, generation relative to the year before repowering in percent on the vertical axis. Before the work the estimates are +3.9, +2.3 and minus 4.5 against a base of 0; the repowering year is minus 6.7; then +33.9, +46.6, +54.7, +59.4 and +61.6 from year 1 to year 5. A shaded 95% interval is drawn from the repowering year onward, the only years for which the memo prints bounds. A dashed line marks the naive regression at +30.4. Headline: +48.4% from the first full year, interval +36.3% to +61.4%.% vs the year before repowering−20020406080−4−3−2−10+1+2+3+4+5Years since repoweringnaive +30.4%-4: +3.9-3: +2.3-2: −4.5-1: base year, 0 by construction0: −6.7, 95% interval −14.2 to +1.5+1: +33.9, 95% interval +20.9 to +48.1+2: +46.6, 95% interval +34.4 to +59.8+3: +54.7, 95% interval +40.2 to +70.6+4: +59.4, 95% interval +43.5 to +77.0+5: +61.6, 95% interval +41.6 to +84.4placebo p = 0.268repoweringyear −6.7+48.4% from year +1interval +36.3% to +61.4%
Repowered plants against never-repowered plants of the same vintage: +48.4% from the first full year after the work, 95% interval +36.3% to +61.4%. Read how it was measured.04 · USGS turbine database, EIA-923
Data behind this chart
Event-time estimates, 04
YearEstimate95% intervalCohortsPlants
−4+3.9not printed780
−3+2.3not printed780
−2−4.5not printed780
−10 (base)not printed780
0−6.7−14.2 to +1.5780
+1+33.9+20.9 to +48.1780
+2+46.6+34.4 to +59.8673
+3+54.7+40.2 to +70.6567
+4+59.4+43.5 to +77.0461
+5+61.6+41.6 to +84.4338
Headline, +1 onward+48.4%+36.3% to +61.4%780
Naive regression+30.4%+23.7% to +37.5%

Measured, dated, re-run

  • 186,229,882 import rows, each joined to the meaning its code carried that month, 1988 to July 2026 05 · Statistics Canada
  • 20,692,840 clickstream events across 4,535,941 sessions, loaded in one glob read 02 · cosmetics shop clickstream
  • 21,890 procurement award and amendment rows, 91 columns, keyed to 13,296 contracts 03 · CanadaBuys
  • 14,077 plant-years of wind generation across 1,486 plants, 2013 to 2025 04 · USGS and EIA-923
  • 213,456 hours of Ontario demand, 2002 to 2026, forecast from 602 daily origins 06 · IESO
  • 919 published figures re-run against fresh pulls, every Monday and on every push ci · GitHub Actions

Every figure the CI re-runs, one mark each

Every figure the CI re-runs, one mark eachUnit chart, one square per published figure, in five columns by project, filled from the baseline up. 01: 56 of 56 reproduced exactly. 02: 22 of 22. 03: 334 of 337, 3 declared. 04: 215 of 220, 5 declared. 06: 277 of 284, 6 drifted inside band and 1 declared. In all 919 figures on the badge of 2026-09-09: 904 reproduced exactly, 6 drifted inside band, 9 declared, none failed.01: 56 of 56 reproduced exactly02: 22 of 22 reproduced exactly03: 334 of 337 reproduced exactly, 3 declared04: 215 of 220 reproduced exactly, 5 declared06: 277 of 284 reproduced exactly, 6 drifted inside band, 1 declared0156 of 560222 of 2203334 of 33704215 of 22006277 of 284
Every figure the CI re-runs, one mark eachUnit chart, one square per published figure, in five columns by project, filled from the baseline up. 01: 56 of 56 reproduced exactly. 02: 22 of 22. 03: 334 of 337, 3 declared. 04: 215 of 220, 5 declared. 06: 277 of 284, 6 drifted inside band and 1 declared. In all 919 figures on the badge of 2026-09-09: 904 reproduced exactly, 6 drifted inside band, 9 declared, none failed.01: 56 of 56 reproduced exactly02: 22 of 22 reproduced exactly03: 334 of 337 reproduced exactly, 3 declared04: 215 of 220 reproduced exactly, 5 declared06: 277 of 284 reproduced exactly, 6 drifted inside band, 1 declared0156of 560222of 2203334of 33704215of 22006277of 284
  • reproduced exactly
  • drift inside band
  • declared
  • failed: 0
One square per published figure, a column per project, filled from the baseline up. On the badge of 2026-09-09, 904 of 919 came back exactly, 6 moved inside their drift bands after the IESO rewrote its 2026 files, and 9 are declared with a reason.ci · GitHub Actions
Data behind this chart
Figures per project on the badge of 2026-09-09
ProjectListedReproducedDrift in bandDeclared
01 Seller risk565600
02 Funnel222200
03 Defence33733403
04 Wind22021505
06 Demand28427761
Total91990469

Every figure on this site traces to a public notebook, and the CI re-runs five of the six projects on every push.How the figures are re-run

One figure, traced

Every figure on this site traces to a public notebook. Here is one, from the file to the number: the day-ahead error of the Ontario demand forecast, project 06.

The day-ahead error, traced from the file to the figure

Ontario hourly demand, the week of 2026-07-13, with the forecast's error by leadLine chart of Ontario hourly demand for the week of 2026-07-13, Monday to Sunday, with a band about the line for the model's mean absolute error by lead over 602 origins, 523 MW a day ahead widening to 1,029 MW a week ahead, and two thin edges for the same-hour-last-week forecast's 1,221 to 1,214 MW. The peak is 25,646 MW at hour 17 on 2026-07-14. The month, July 2026, is drawn as a strip below.15,00020,00025,000 MWMonTueWedThuFriSatSunorigin, midnight2026-07-14, hour 1725,646 MW523 MWa week ahead, model 1,029 MWsame hour last week 1,214 MW
Ontario hourly demand, the week of 2026-07-13, with the forecast's error by leadLine chart of Ontario hourly demand for the week of 2026-07-13, Monday to Sunday, with a band about the line for the model's mean absolute error by lead over 602 origins, 523 MW a day ahead widening to 1,029 MW a week ahead, and two thin edges for the same-hour-last-week forecast's 1,221 to 1,214 MW. The peak is 25,646 MW at hour 17 on 2026-07-14. The month, July 2026, is drawn as a strip below.15,00020,00025,000 MWMonTueWedThuFriSatSunorigin, midnight2026-07-14, hour 1725,646 MW523 MWa week ahead, model 1,029 MWsame hour last week 1,214 MW

July 2026, hour by hourthe week of 2026-07-13 above

523 MWday-ahead error of a linear model with no weather input, 3.0% of demand, against 1,221 MW for same hour last weekSee 06

The week of 2026-07-13, hour by hour, from the month in the strip. The bands are the model’s and the same-hour-last-week forecast’s mean absolute error by lead over the 602 origins, drawn about the actual demand: the average miss, not this week’s forecast. The bracket measures the day-ahead figure.06 · IESO
Data behind this chart
What this chart prints, and where its line comes from
Peak of the week25,646 MW at hour 17 on 2026-07-14
Model, mean absolute error by lead523 MW a day ahead; 780 two days ahead; 954 days 3 to 6; 1,029 a week ahead
Same hour last week, by the same leads1,221; 1,221; 1,219; 1,214 MW
Origins scored602, every day from 2024-12-31 to 2026-08-24
Hours in the series213,456, 2002 to 2026
The line and the stripOntario demand, hour ending 1 to 24 of each day of July 2026, from the IESO's published file for 2026 as the notebook loads it. The values are in that file and in the notebook's tables, not in the memo, and the chart draws them without printing them.
  1. The file.213,456 rows, one per hour from 2002 to 2026, as the IESO publishes them: a date, an hour, a demand in megawatts. One month of them is here, July 2026.
  2. The series.Read in order, the rows are a wave. It crests every afternoon, drops every weekend, and on 2026-07-14 a heat wave lifts it to 25,646 MW at hour 17.
  3. The matrix.The notebook folds the series one row per day and one column per hour. Every day from 2024-12-31 to 2026-08-24 is an origin, 602 of them. At each midnight, a forecast for the next 168 hours from sixteen things known at the origin, and no weather.
  4. The finding.A day ahead, the model misses by 523 MW on average, 3.0% of demand. Same hour last week misses by 1,221 MW. By the seventh day it is 1,029 against 1,214, and the two can barely be told apart. The figure in the ledger below is this one.

Work

Six analyses and the CI that checks them. Each page states the question, the defects found in the source before any analysis, the method, the finding, and what the data cannot say.

Experience

About three and a half years in analyst roles, the first at a training organization in Hyderabad, the second at an early-stage manufacturer in Vaughan, Ontario.

Pfaff Technologies

Strategy Analyst, Jan 2026 to Jul 2026

Business Development Analyst, Sep 2025 to Jan 2026

Vaughan, ON · about 10 months

One generalist role at an early-stage carbon-fibre composites and additive-manufacturing startup, re-levelled once. The company wound down in July 2026.

  • Owned the market-entry research and evaluation program across three industries, medical, aerospace and defence: domain and regulatory research, market sizing, competitor mapping, business cases with feasibility and ROI analysis, scenario forecasts, and a go/no-go recommendation to leadership for each opportunity. The opportunities were assessed, not entered.
  • Built the comparison and scoring models in Excel: weighted scoring matrices with explicit criteria weights, best, base and worst scenario toggles driven from an input cell, what-if data tables, and pivots with XLOOKUP and INDEX-MATCH.
  • Three lines of that program, carried through to numbers on public data:
    • Aerospace, wind-turbine bladesAt Pfaff: market sizing, competitor mapping, the business case and the go/no-go recommendation, in reports to leadership. Then, on USGS and EIA data: what repowering does to a plant’s generation, measured on 80 repowered plants against 498 never-repowered plants of the same vintage across 14,077 plant-years, +48.4% from the first full year after the work. See 04
    • Defence, tactical low-cost dronesAt Pfaff: market and competitive intelligence, with the regulatory and Controlled Goods requirements, in reports to leadership. Then, on the CanadaBuys award file: what the government’s own data says about the 70% Canadian-firm target under each definition, who the incumbents are at legal-entity and parent grain, and that uncrewed and autonomous systems are $2.1M of the $10.9B in classified defence value. See 03
    • Additive manufacturingAt Pfaff: market sizing and competitive intelligence for the company’s own 3D-printing capability, in reports to leadership. Then, on Statistics Canada import files: the import series for 3D-printing machines rebuilt across the January 2022 tariff change, 186 million rows through a dbt Core project on DuckDB. See 05
  • Built and owned the ISO 9001:2015 quality-system planning and documentation from scratch, documents, templates and workflow plans, and selected the software running the program. Implementation was underway when I left; certification had not yet been achieved.
  • Authored the company’s full Controlled Goods Program security plan for its designation audit under the Defence Production Act: safeguarding, personnel security, access control, IT security, training and breach reporting. The audit passed and the company’s designation was approved.
  • Contributed technical research and project coordination to a DND low-cost-drone challenge entry on which Forward Robotics was the primary applicant and Pfaff the manufacturing partner. The entry won round one.
  • Built and managed marketing and SEO dashboards in Google Analytics with Google Tag Manager, and business-performance reports and dashboards in Power BI and Tableau.
  • Researched and identified federal funding programs, NRC IRAP, NGen, CanExport and FedDev Ontario; drew cross-functional process maps in Miro; selected, configured and administered ClickUp.

Karting Operations Platform

Apr to Jun 2026 · inside the Strategy Analyst period

For Mosport Karting Centre, an affiliated entity under the same parent. I architected, directed and reviewed a production system; an AI coding agent wrote most of the code under my direction and gating. Discovery began with a 5,378-row booking audit I ran by hand in Excel, pivots, cross-sheet lookups and rule-based deduplication under matching rules I defined, which found about 30% duplicates. I designed the Postgres data model, versioned immutable rule-sets with deny-by-default row-level security on every table, and engineered a PIPEDA and COPPA compliance posture for about 695 member records held in Canada. Race weight-class classification, a critical function, runs as two implementations held identical by an 11-fixture parity gate, and the cutover replay against the legacy system closed with 56 diffs and 0 unadjudicated. Class thresholds were inferred at about 80% match and have not been validated.

QAANS

Business and Operations Analyst, Jan 2021 to Jul 2023

Hyderabad, India · 2 years 7 months, alongside the BEng

  • Built and delivered the organization’s regular reporting in Excel and Power BI for the executive team: revenue, KPIs, churn, course-level performance, enrolment metrics, learner demographics and marketing performance.
  • Oversaw day-to-day operations and marketing for a training organization: bookings tracked and managed in Excel, reporting, and marketing performance by channel, including lead volume, cost per lead and per acquisition, stage-by-stage funnel conversion, and campaign ROI and ROAS.

Skills, and where each is evidenced

A claim is only as good as the place you can check it. Project numbers refer to the work above; the rest is employment.

SkillEvidence
Analysis
Analytical SQL on raw event, transaction and trade data in DuckDB: typed staging from all-text loads, keyed dedup with ROW_NUMBER and QUALIFY, LAG diagnostics, running shares, list functions over multi-valued fields, a temporal join on a code and a month against validity ranges, assertion suites with expected-against-found counts01 to 06
Python for analysis: pandas, NumPy, SciPy, statsmodels, matplotlib; hand-written estimators and bootstraps01 to 06 and ci; postgraduate diploma
Quasi-experimental analysis: staggered adoption, a vintage-matched control pool, pre-trend checks, a cohort-by-year estimator written by hand and cross-checked to six decimals, Goodman-Bacon decomposition, minimum detectable effect and power04
Forecasting with an honest baseline: rolling-origin backtest, seasonal-naive baseline, one regression per lead, Diebold-Mariano with an overlap correction, empirical intervals judged by coverage and sharpness06
Break tests and series reconstruction: Chow test, rolling Chow scan, interrupted time series with Newey-West errors, moving-block bootstrap05
Definition-sensitivity and spend analysis: readings on two scopes and two measures, entity resolution through published crosswalks, HHI and top-share at entity and parent grain03
Data engineering and reproducibility
dbt Core on DuckDB: sources over hive-partitioned Parquet, staging views, a validity-range dimension, an incremental fact by year, seeds, 71 tests, docs and lineage from the manifest05
Data contracts, source-to-target maps, decision logs, pitfalls records, published crosswalks03, 04, 05, 06
Star schemas exported and reconciled row for row, with a DAX measure layer and a row-level-security role specified; the dashboards themselves are not yet built03, 04, 05, 06
Reproducibility engineering: a numbers manifest per project, headless notebook execution, source fingerprinting, a drift contract, GitHub Actions with a matrix and an orphan reports branchci
PostgreSQL data modelling: versioned immutable rule-sets, constraints, triggers, deny-by-default row-level securityKarting Operations Platform; the design decisions were mine
Reporting and spreadsheets
Excel, advanced: weighted scoring matrices, scenario toggles, sensitivity tables, pivots with XLOOKUP and INDEX-MATCHPfaff models; the 5,378-row booking audit; QAANS reporting
Power BI and Tableau reportingPfaff; Power BI at QAANS, Jan 2021 to Jul 2023
Google Analytics with Google Tag ManagerPfaff
Strategy, quality and process
Market and competitive intelligence, business cases, go/no-go recommendationsPfaff
Quality and compliance documentation: ISO 9001:2015 QMS, Controlled Goods security planPfaff
Project and process: CAPM; ClickUp administration; process mapping in Miro and VisioPfaff; certification

Education

Mechanical engineering first, then a postgraduate diploma in data analytics after moving to Canada in 2023.

Degrees and diplomas

  • Postgraduate Diploma, Data Analytics for BusinessSt. Clair College, OntarioSep 2023 to Apr 2025
  • BEng, Mechanical EngineeringMuffakham Jah College of Engineering and Technology, Hyderabad2019 to 2023
  • BA (Hons), PsychologyIndira Gandhi National Open University, distance mode2020 to 2023

Certifications

  • CAPM, Certified Associate in Project ManagementPMI
  • Google Data Analytics CertificateGoogle, Coursera

Contact

I am looking for a data, BI, business or strategy analyst role in the Greater Toronto Area. Email is the fastest way to reach me. If a number on this site interests you, the page it sits on links to the notebook that produced it.

Authorized to work in Canada (PGWP), no sponsorship required.