Waqar Ahmed

AI Workflow Automation (n8n & Make)

Your repetitive hand-offs between apps — leads arriving, records updating, people being notified — built into an automatic workflow that runs the same way every time, with duplicates stopped before they ever reach your sheet or CRM.

What you get

• A working automation built in n8n or Make, connected to your tools
• De-duplication and validation built in, so bad or repeated data never passes silently
• Tested end-to-end with batch data before handover, with the run results shown to you
• A clear runbook: what it does, what it watches, and what to check if it stops

How it works

You describe the hand-off in writing and share access to the tools involved. I build it, test it on real-shaped sample batches, show you the results, then switch it on with you.

Verified on a lead-intake build run live in both tools: batch of 8 submissions → 6 added, 2 duplicates skipped; second batch of 5 → 2 added, 3 skipped — every count exact.

AI Chatbot / FAQ Bot for Your Website

A chatbot that answers your customers' questions from your own approved business information — your services, prices, policies and FAQs — and says so honestly instead of guessing when a question falls outside what you provided.

What you get

• A working bot built on a structured knowledge base of your approved content (the verified sample used a 45-entry knowledge base)
• A written instruction block defining its identity, scope and hard limits — it never invents prices, policies or facts
• A full test report: every test question, the bot's answer, and the check against your source material
• Deployment where your customers already are — your website, inbox or social account

How it works

You supply the information the bot may use. I build the knowledge base, write its rules, test it question-by-question — including questions designed to catch it guessing — and hand over the test report with the bot.

BI Dashboards (Looker Studio)

Your spreadsheet or sales data turned into a live business dashboard — revenue, margins, trends and breakdowns that refresh from your data instead of being re-typed into slides every month.

What you get

• A live Looker Studio dashboard built on your data (the verified sample build: 4 pages, 8 data sources)
• Every headline figure validated one by one against the source data before delivery
• Filters and breakdowns your team can actually use — by period, product, channel or region
• A short written guide to reading and refreshing the dashboard

How it works

You share the data file or sheet. I clean and model it, build the dashboard, then validate every figure on it against the source and show you the validation results.

CRM Migration & RevOps Data Services

Moving between CRMs — or rescuing the data inside the one you have — starts with the data itself: cleaned, de-duplicated, standardised and shaped to the target system's fields, so the import lands clean instead of importing years of mess.

What you get

• An import-ready contact file for the target CRM (verified sample: 1,000 exported contacts → 905 import-ready)
• Duplicates caught three ways — exact, fuzzy on email, fuzzy on phone + name — and merged safely
• Phones standardised to international (E.164) format, addresses and company names normalised, dates standardised
• A separate review file for every record that needs a human decision, with the reason stated (110 rows in the sample) — nothing silently dropped: 905 + 95 excluded = the full 1,000, reconciled exactly

How it works

You export your current CRM data and share it. I run the cleanup and shaping, reconcile every record in and out, and deliver the import-ready file plus the review file and validation report.

Technical Documentation & SaaS Help Centres

Documentation your users and support team can actually rely on: API references, getting-started guides and step-by-step SOPs, written to a defined style sheet and mechanically cross-checked so every example, field name and error code agrees across every page.

What you get

• A style sheet that fixes terminology, formatting and tone for the whole set
• API reference documentation (verified sample: 10 endpoints, 8 object tables, 8 error codes — every use cross-verified)
• A step-by-step getting-started guide (6 steps in the sample)
• Operational SOPs (sample: a failed-webhook-delivery procedure, 12 steps with a 3-level escalation path)
• Every code example parsed and checked — in the sample, all 17 JSON examples valid, all invoice arithmetic exact

How it works

You share your product details, existing docs or API spec. I agree the style sheet first, then write the set and run mechanical cross-checks over the whole thing before delivery.

AI-Assisted Micro-Tool Builds (fixed scope)

One small, sharp tool that does one job perfectly — like the verified sample, a CSV cleaner that runs entirely in a single HTML file: drop in a messy CSV, get a cleaned file and a report of everything that was fixed and flagged.

What you get

• A single-file tool that works offline in the browser — no installation, no server, nothing to maintain
• A live demo run on sample data, with counts you can check (sample: 20 rows in → 16 out; 14 dates standardised across 5 formats; 3 invalid emails flagged)
• A README stating exactly what the tool does — and its honest limits
• The tool is genuinely reusable: the sample CSV Cleaner is one I use in my own client spreadsheet work

How it works

You describe the one job in writing. I agree the fixed scope, build the tool, and verify its output against an independent re-calculation before delivery.

EUDR Supplier Traceability Data Preparation

The EU Deforestation Regulation requires due-diligence data on suppliers, production plots and shipments. I prepare that data — collected, cleaned, geolocation-checked and structured into due-diligence-ready files — for commodity supply chains such as cocoa and coffee.

What you get

• A supplier-data pack on your chain (verified sample: 15 suppliers, 45 plots as GeoJSON, 23 consignments)
• Every planted error class in the sample caught and classified — missing geolocation, invalid coordinates, duplicate suppliers, broken consignment links and more
• Files structured for your compliance team or DDS tooling to pick up directly
• Timing that respects the regulation: it applies from 30 Dec 2026 for large/medium operators and downstream, and 30 Jun 2027 for micro/small

How it works

You share your supplier, plot and shipment records. I validate, clean and structure them, flag every gap with its reason, and deliver the pack with a validation report.

Financial Modelling — Data Work (your assumptions, exact mechanics)

A three-statement financial model built or audited purely as a data machine: every calculation a live formula linked to your assumptions, balance sheet balancing to the cent, cash tying out, and an audit that finds hard-coded plugs and broken links wherever someone hid them.

What you get

• A 3-statement model (sample: FY2024 actuals + FY2025–27 forecast) with 188 of 188 calculation cells as formulas linked to the inputs — no hard-coded answers
• An audit report: the clean model passed 9/9 structural checks (balance sheet balances in all 4 periods, cash tie-out, retained-earnings roll-forward, debt and depreciation ties)
• Break-finding proven: the audit caught the sample's 2 deliberately planted breaks exactly, with their knock-on amounts
• An independent rebuild from your assumptions file, diffed figure-by-figure against the model

How it works

You provide the assumptions — they are always yours. I build or audit the mechanics, run the structural checks, and report every break found with its size.

Airtable & Notion Workspace Builds

A working operating system for a small team, designed and delivered as a complete build pack: an Airtable base and a Notion workspace covering the same business — clients, projects, tasks and invoices — with every relationship, formula and total checked.

What you get

• Airtable: full schema (sample: 4 tables, 12 views, formulas), ready-to-import CSVs, and an import guide
• Notion: workspace blueprint (4 databases, relations, 3 page templates), matching CSVs, and a setup guide
• One consistent dataset across both (sample: 8 clients, 12 projects, 30 tasks, 15 invoices — 65 records identical field-by-field in both packs)
• Totals independently recomputed: billed 27,850 · paid 23,150 · outstanding 4,700 in the sample data

How it works

You describe how your team works in writing. I design the schema, build the pack, verify every link and total, and deliver it with step-by-step import/setup guides.

Bookkeeping Cleanup — Support for QuickBooks / Xero Users

Months of messy transactions turned into a reconciled workbook: bank and books cleaned side by side, duplicates removed, every difference hunted down and listed with its reason — so your bookkeeper or accountant starts from numbers that tie to the cent.

What you get

• A cleanup workbook: cleaned bank sheet, cleaned books sheet, a live-formula reconciliation, and an exceptions list
• Sample result (3 months, fictional business): bank 90 lines and books 85 records cleaned; every planted mess caught exactly — duplicates 4 + 6, entries missing from books 8, bank-only fees/interest 5, outstanding items 6, digit-transposition typos 5, wrong dates 5, payee name variants 12, wrong categories 7
• Reconciliation tying exactly to the bank's closing balance — difference $0.00 — with 73 matched pairs and 44 exceptions, each with a stated reason and next step
• Personal-looking expenses flagged for your decision — never judged, never quietly reclassified (5 flagged in the sample)

How it works

You share the bank export and the books export. I clean, match and reconcile them, account for every raw row exactly once, and deliver the workbook plus a plain-language cleanup report.

Podcast & YouTube Written Repurposing Pack

One episode of your podcast or channel turned into a complete written pack — show notes, chapters and ready-to-post social content — built only from your episode's own words. Nothing invented, nothing padded: every quote is verbatim and every number traces back to the episode itself.

What you get

• Show notes: 3 title options, a short and a long summary, key takeaways and topics list
• YouTube chapters in standard format, timed proportionally from the narration script and labelled as estimates
• Social posts written for each platform: a LinkedIn post, an X thread (every post within the 280-character limit), an Instagram/Threads caption and a Facebook post
• Verbatim quotable moments, an SEO keyword list drawn from the episode's own terms, and a short blog/newsletter version
• Sample pack, from our own EIDOS episode: 5 chapters, 6-post X thread, 6 verbatim quotes, 10 keywords, 377-word blog version — numbers audit clean

How it works

You share the episode's script or transcript. I extract, structure and write the pack, then verify every quote and number against your source before delivery.