Client engagements
Consulting delivery
Exhibit 1A float-glass manufacturer2026
Float glass market assessment
Engagement
Market assessment of Iraq's float glass sector for a manufacturer weighing a local position.
My role
Owned the reissue. Rebuilt the delivered deck to carry every correction from two rounds of client clarification, edited inside the client's own file so the design stayed intact.
What it gave the client
The client got a deck it could circulate internally without a covering note, with the limits of the market estimate stated on the pages that used it.
30pp
Corrected deck reissued in the client's own template
24
Written commitments audited before release
Market sizingTrade dataClient deliverables
Source: Two client clarification rounds; independent audit against all written commitments, 2026.
Exhibit 2An industrial investor2025 – 2026
Paper sack plant feasibility study
Engagement
Feasibility study for a 100 million bag per year multi-wall paper sack plant serving the cement sector.
My role
Built the demand model and the supply-side map, and set the triangulation approach across cement plants, domestic producers and shipment-level import data.
What it gave the client
The investor could size the opportunity against real capacity rather than headline cement tonnage, and saw the structural risk in the bagged-to-bulk shift before committing capital.
~28
Cement plants covered, across four independent sources
4
Domestic producers mapped by capacity
Feasibility studyDemand sizingTrade dataCost modelling
Source: Four expert sources across roughly 28 cement plants; shipment-level import data; producer capacity mapping, 2025–26.
Exhibit 3A Japanese automaker's national distributor2025 – 2026
Nationwide brand health assessment
Engagement
Brand health assessment across the Iraqi passenger car market, built on Aaker's brand equity model.
My role
Worked on the engagement team through instrument design, fieldwork and analysis, across a national survey, the interview programme and mystery shopping.
What it gave the client
The distributor got a measured read on where its position was eroding, rather than the sales figures it already had.
17
Brands measured through the equity funnel
27
Key-informant interviews across the trade
Brand equityAaker modelMystery shoppingTrilingual fieldwork
Source: National survey against a quota design, 27 key-informant interviews, mystery shopping across seven or more cities, 2025–26.
Exhibit 4A global automaker and its Iraqi dealer2025
Market and aftersales study
Engagement
A 14-week market and aftersales study supporting a relaunched dealer network, subcontracted from a regional consultancy.
My role
Worked on the team that owned all primary data collection inside Iraq. The prime contractor held desk research and deck production.
What it gave the client
The client got ground-level aftersales, parts pricing and dealer data that could not be gathered from outside the country.
14 wks
Original delivery window
100%
Of in-country primary fieldwork on our side
Aftersales researchParts pricingSubcontracted fieldwork
Source: Multi-method market and aftersales fieldwork across Iraq, 2025.
Exhibit 5A prospective distributor for a European automaker2025 – 2026
Competitive and pricing study
Engagement
Competitive assessment of a European marque's focus models against rising Chinese brands, in a market where imported nearly-new cars compete with new sales.
My role
Worked on the engagement team across consumer interviews, mystery shopping and the price mastersheet build.
What it gave the client
The client could compare its models against rivals at both authorised and parallel-import prices, which is not available from any published source.
324
Model-trim rows priced across both channels
21
Outlets mystery shopped
24
Consumer in-depth interviews
Competitive benchmarkingPrice researchMystery shopping
Source: 24 consumer in-depth interviews across two cities, 21 mystery-shopped outlets, 324-row dual-channel price mastersheet, 2025–26.
Exhibit 6A United Nations agency2026
Housing affordability study
Engagement
Housing affordability indicators for a major Iraqi city, feeding the agency's master plan work.
My role
Designed the sampling approach, selected and verified the imputation method for missing survey income, and built the indicator set with confidence intervals.
What it gave the client
The agency received indicators at neighbourhood level with their limitations stated, which its own analysts could re-run through a prototype application.
35
Sample clusters on master-plan geographies
4,353
Missing survey values imputed and independently verified
Survey methodologyComposite indicatorsImputation
Source: Commissioned household survey across 35 clusters; supplementary listing collection, 2026.
Exhibit 7An enterprise technology company2025
Market assessment, two service lines
Engagement
Market assessment for managed document services and cybersecurity in Iraq.
My role
Worked on the engagement team across interview design and analysis, with criteria weighted so willingness to invest was separated from stated interest.
What it gave the client
The client saw where real budget existed rather than where interest was expressed, in a category most of the market could not yet name.
25
Organisations interviewed
B2B researchKey-informant interviewsCategory sizing
Source: 25 organisations interviewed across more than ten sectors, 2025.
Exhibit 8A new-entrant internet services provider2025
B2B demand assessment
Engagement
Demand assessment for business internet, hosting, cloud and security services ahead of a market entry.
My role
Worked on the team that ran a two-sided interview programme, reading supply-side and demand-side accounts against each other.
What it gave the client
The client entered with a view of what businesses would actually buy, in a category where a stated-preference survey would have misled it.
44
Interview transcripts across both sides of the market
10
Demand-side sectors covered
B2B demand assessmentTwo-sided research designMarket entry
Source: 16 supply-side and 28 demand-side key-informant interviews across ten sectors, 2025.
Exhibit 9A telecom operator in the Kurdistan Region2025 – 2026
Segmentation and loyalty study
Engagement
Youth market study covering behavioural segmentation, churn drivers, dealer network performance and loyalty design.
My role
Worked on the engagement team, running the clustering and carrying the segments through into the offer concepts.
What it gave the client
The operator got segments it could act on commercially, with loyalty and offer designs already built against them.
5
Behavioural segments carried into offer design
61 + 408
Field transcripts and survey responses
SegmentationChurn analysisNPSLoyalty design
Source: Nine engagement stages, 61 field transcripts and 408 survey responses, 2025–26.
Exhibit 10An online travel platform2026
Market entry evidence report
Engagement
A lead-generation report answering a foreign party's standing objections to entering Iraq.
My role
Built the report end to end, from research design through an adversarial review round to the final 50-page document.
What it gave the client
The client had a sourced document it could hand to a hesitant counterparty, anchored on foreign companies already operating in the country.
~95
Named foreign entrants verified as evidence
50pp
Final designed report
Market entryThought leadershipReport design
Source: Desk research with adversarial senior-analyst review; roughly 95 verified foreign entrants, 2026.
Exhibit 11A US-funded civic education programme2026
External programme evaluation
Engagement
External evaluation of a school-based civic engagement programme against a 19-indicator framework.
My role
Data analyst on the evaluation team, responsible for reconciling the data streams and re-deriving the reported figures before submission.
What it gave the client
The donor received a report whose numbers had been independently reproduced rather than carried forward from the programme's own reporting.
19
Indicators in the evaluation framework
Monitoring & evaluationDonor reportingQualitative coding
Source: Nine data streams scored against a 19-indicator framework, 2026.
Data products and applied AI
Built, validated, shipped
Exhibit 12KAPITA Research, public platform2026 – ongoing
iData, Iraq's public statistics portal
Engagement
An open platform publishing Iraqi government statistics that are otherwise hard to use.
My role
Product manager. The data architecture, the taxonomy and the definition of what a dashboard is are mine to decide and defend.
What it gave the client
A full audit of the estate found published dashboards that no reader could reach through browse or search. The rebuild moves the platform onto an indicator model with a publish check that blocks incomplete records.
180
Prepared dashboards audited
571
Live charts under review
23
Published dashboards found unreachable
Product architectureInformation designData modellingBilingual AR / EN
Source: Full-estate census of 42 pipelines, 180 prepared dashboards and 571 live charts, 2026.
Exhibit 13A commercial bank in Iraq2026
Arabic financial statement reader
Engagement
An automated reader turning scanned Arabic year-end accounts into the bank's own credit analysis workbook.
My role
Designed, built and validated the system, including the checking layer that decides which figures a human must review.
What it gave the client
Credit officers stopped re-keying roughly 78 fields per file by hand, and every figure the system was unsure about was escalated rather than passed through. The bank bought it at board level.
~90%
Line-item accuracy on real client documents
~700
Figures reviewed with no silent error surviving
Document AIHuman-in-the-loop designArabic OCROn-premise deployment
Source: Client document set; independent validation against 20 unseen scans from a public filings corpus, 2026.
Exhibit 14KAPITA Research, internal product2025 – 2026
Sector mapping engine
Engagement
A reusable pipeline that maps and sizes a consumer sector at branch level, and writes the report.
My role
Designed and built it, and set the classification rules that have to be encoded before a new sector can run.
What it gave the client
The firm stopped rebuilding each sector study from scratch and now sells the output as a repeatable product.
2
Consumer sectors shipped to date
29pp
Report generated per run, alongside a workbook
ProductisationMarket sizingPipeline design
Source: Sector runs across private K-12 schools and gyms, 2025–26.
Exhibit 15Investor-facing, KAPITA Research2026
Iraq economy flow model
Engagement
A visual model of how money moves through the Iraqi economy, built for investors who read GDP per capita and conclude there is no market.
My role
Built the model and the visualisation, including the source tiering that labels which figures are estimates.
What it gave the client
Investors get the distribution rather than the average, with every node stating how many people it covers and how much each receives.
16
Nodes in the flow model
Data visualisationEconomic structureSource tiering
Source: 16-node flow model on tiered sources; estimate-tier figures labelled on the chart, 2026.
Exhibit 16Independent build2026
Diabetes companion, clinical safety design
Engagement
An Arabic conversational companion helping low-income patients manage insulin between clinical visits.
My role
Built it as a physician, and designed the safety layers that sit outside the model so dosing requests are refused at the routing layer.
What it gave the client
Patients get logging, pattern warnings and education from a system that cannot be talked into giving a dose.
3
Independent safety layers outside the model
60
Unit and live safety tests
Clinical safety designConversational AISwiftUIArabic NLP
Source: 48 unit tests and 12 live safety smokes, 2026.
Exhibit 17A medical-travel company2026
Customer agent and sales dashboard
Engagement
An Arabic sales agent handling live social-channel enquiries, with a dashboard behind it.
My role
Run the whole operation: intake sequence, escalation rules, the knowledge base the owner updates by voice, and the commercial reporting.
What it gave the client
The company answers enquiries around the clock in Iraqi dialect, hands over to a human when the agent should not proceed, and can see cost and margin per deal.
~1.5¢
Cost per reply, logged per message
4
Dashboard views from funnel to escalation
Conversational AIIraqi ArabicCRM designUnit economics
Source: Live customer conversations with per-message cost logging, 2026.
Large-scale data collection
Where the dataset had to be built first
Exhibit 18A global beverage company2026
Premium outlet census
Engagement
Identification of the premium food-service outlets in Baghdad worth a direct commercial approach, where no register exists.
My role
Ran the census. Built the automated discovery, then designed the manual verification pass and the evidence record behind every inclusion and removal.
What it gave the client
The client got a call list it could audit, with the reason for each exclusion recorded and an interactive page letting it move the filters and watch the count change.
2,796
Candidate venues screened
161
Verified outlets delivered
65
Carrying confirmed price evidence
Sector censusEvidence designClient dashboards
Source: Automated discovery with per-outlet manual verification and captured evidence, 2026.
Exhibit 19An importer of fragrance concentrates2026
Retail census and route planning
Engagement
A census of the capital's fragrance retailers ahead of the client moving from wholesale to direct supply.
My role
Ran the collection and the classification, judging every shop that carried photographic evidence rather than sampling.
What it gave the client
The client got a visit universe with contact details, and a measured count of shops with no digital trace at all, which is the argument for funding physical fieldwork.
2,020
Retail locations mapped
829
Photo-carrying shops judged individually
918
Phone numbers recovered
Retail censusImage classificationTerritory analysis
Source: 2,020 mapped locations; every photo-carrying shop judged individually, 2026.
Exhibit 20A global beverage company2026
Field team route planner
Engagement
A planner turning a target list into daily routes for a field sales team.
My role
Built the optimiser and the interface, including the choice of objectives and the measured trade-off shown against each.
What it gave the client
The client can pick between shortest distance, highest value first, or one district per day, and see what each choice costs before committing to it.
6–36%
Measured reduction in travel distance
60
Configurations tested against the baseline
Operations researchGeospatialDecision tooling
Source: 60 tested configurations against a nearest-neighbour baseline, 2026.
Capability
What I actually do
Research
Survey design and
sampling, cluster design, composite indicators with confidence intervals, imputation method
selection, segmentation and clustering, qualitative fieldwork and key-informant interviews,
mystery shopping, competitive benchmarking, market sizing, monitoring and evaluation of donor
programmes.
Data products
Product ownership of
a public statistics platform — data architecture, taxonomy, publish contracts and validation,
information design for charts and indicators, bilingual Arabic and English delivery.
Applied AI
Document extraction
pipelines, human-in-the-loop validation design, deterministic safety layers for clinical contexts,
evaluation harnesses and gold-set construction, on-premise open-weights deployment planning,
large-scale web and social data collection.
Build
R, Python, SQL; Cloudflare
Workers and D1; SwiftUI; interactive dashboards and decision tools. System design and validation
strategy are mine; implementation is substantially machine-assisted and checked.
Delivery
Board, ministry and
embassy presentation; consulting decks and structured analytical reports; methodology defence to
sceptical clients; cross-functional team leadership.
Sectors
Banking and finance,
telecom, automotive, healthcare, industrial manufacturing, FMCG, retail, real estate, and the
donor and development sector.
Languages
Arabic (native)
· English (C1) · French (B1) · German (in progress)
