Portfolio

Olusegun Marcus

Operations professional turned automation builder. I design AI-assisted workflows with Make.com, n8n, Google Workspace and Gemini, and test them against real data, including where they fail.

Lagos, Nigeria · Available for remote roles worldwide

About

My background is in operations and business management, so I build workflows around how work actually gets done. I have been studying data science and automation on my own since 2022.

My method is AI-assisted workflow design: I plan the process, use AI tools to build and debug, then test every step myself so I can explain it.

Projects

Five automations, tested and documented

Case study

Receipt Data Extraction

Reads receipt photos with Gemini and is scored against public ground-truth labels. On 15 unseen receipts: dates and totals 15/15, company names 3/15.

Read the case study  ·  Download PDF

Case study

Customer Review Intelligence

Reads customer reviews in n8n, asks Gemini for sentiment, theme and urgency, and is scored against human labels. On 100 reviews, sentiment matched 97/100. Themes were acceptable on 24 of 33 hand-checked reviews.

Read the case study  ·  Download PDF

Case study

Support Ticket Triage

Sorts banking support tickets into 10 categories with n8n and Gemini, scored against the dataset’s labels. On 100 tickets, 96 matched the label. The four misses are listed.

Read the case study  ·  Download PDF

Automation

Remote Job Radar

Fetches remote job listings, scores them with Gemini against a profile, removes duplicates, saves strong matches to Google Sheets and sends Telegram alerts.

See how it works

Automation

Lead Generation Pipeline

Pulls Lagos business listings, scores them as sales leads, removes national chains, saves Hot and Warm leads to Google Sheets and sends a weekly Telegram summary. Built in n8n on free tools.

See how it works

Case study

Receipt Data Extraction with Make.com and Gemini

A workflow that reads photos of receipts and writes the company name, date and total into a Google Sheet. I scored its output against the ground-truth labels of the public SROIE dataset.

Download the full case study (PDF)

Result on 15 receipts it had never seen

15/15
Dates correct
15/15
Totals correct
3/15
Company names correct. This is the failure I found.

All three tests

TestReceiptsCompanyDateTotalOverall
Prompt v1198/1919/1919/1946/57 (80.7%)
Prompt v2, same 19 receipts (re-test)1915/1919/1918/1952/57 (91.2%)
Prompt v2, unseen receipts153/1515/1515/1533/45 (73.3%)

How it works

  1. Google Drive: search a folder of receipt images and download each one.
  2. Gemini: read the image and return company, date and total as separate fields.
  3. Google Sheets: add one row per receipt.
  4. Scoring: answer keys sit on a separate tab and are matched by file name. Each field scores 1 or 0, with strict matching.
The Make scenario for the receipt extractor: Google Drive search, Google Drive download, Gemini upload a file, Gemini extract structured data, then Google Sheets add a row.
The Make scenario. Google Drive finds and downloads each receipt image, Gemini reads it and extracts the company, date and total, and Google Sheets adds a row.

What I found

Evidence from the scoring sheets

These are screenshots of the real scoring sheets, first rows only. In each one, 1 means the AI’s answer matched the answer key and 0 means it did not. The v1 sheet also holds three extra receipts in rows 2 to 4 that are not counted, so its accuracy covers the 19 receipts in rows 5 to 23. The v2 sheet has one receipt that was not in v1 (row 4), which is also left out of its accuracy.

Google Sheet for prompt v1 with columns for the AI’s company, date and total, the true values, match scores of 0 or 1, and an accuracy of 80.70 percent for 19 receipts.
Test 1, prompt v1 on 19 receipts. Company names miss often, while dates and totals mostly match. Overall accuracy: 80.7%.
Google Sheet for the prompt v2 re-test on the 19 shared receipts, with match scores of 0 or 1 and an accuracy of 91.23 percent.
Test 2, prompt v2 re-tested on the same 19 receipts. Most company names now match. Overall accuracy: 91.2%.
Google Sheet for the unseen receipts test. The AI company column repeats MAKASSAR FRESH MARKET SDN BHD where the true company is GARDENIA BAKERIES (KL) SDN BHD, with company match 0 and date and total match 1. Accuracy 73.33 percent.
Test 3, prompt v2 on unseen receipts. Dates and totals match, but the AI repeatedly returned MAKASSAR FRESH MARKET SDN BHD where the answer key says GARDENIA BAKERIES (KL) SDN BHD. Overall accuracy: 73.3%.

Limitations

Case study

Customer Review Intelligence with n8n and Gemini

An n8n workflow that reads customer reviews from a Google Sheet, asks Gemini for a sentiment, a theme and an urgency flag, and writes the results back. I scored it against the human labels in the public UCI Sentiment Labelled Sentences dataset (Amazon reviews).

Download the full case study (PDF)

Result on a 100-review sample

97/100
Sentiment matched the human label
24/33
Themes acceptable in my hand-check of 33 reviews. This is the weak point.
30/33
Urgent flag acceptable against my own written rule

What was checked, and how

MeasureResultHow it was checked
Sentiment (positive or negative)97/100Human labels in the dataset
Positive reviews identified41/43Human labels
Negative reviews identified56/57Human labels
Theme acceptable24/33My manual check, using a rule I wrote beforehand
Urgent flag acceptable30/33My manual check, using a rule I wrote beforehand

How it works

  1. Read the 100 sampled reviews from a Google Sheet (every 10th review of the dataset’s Amazon file: 43 positive, 57 negative).
  2. Send them to Gemini in batches of 5, one request per review, with a 25-second pause between batches to stay inside the free rate limit.
  3. Gemini returns JSON: sentiment, one of eight themes, and an urgent true or false.
  4. Attach the review ID and the human label, and calculate a 1 or 0 for each sentiment answer. A “neutral” answer counts as a miss, because the dataset has no neutral class.
  5. Append one row per review to a results tab. Scoring formulas on a summary tab check that there are 100 rows, 100 unique IDs and no answers outside the allowed values.
The n8n canvas for the review workflow after the full run. Loop Over Items ran 21 times and its done output carries 100 items. Message a model, Edit Fields, Append row in sheet and Wait each ran 20 times.
The n8n workflow after the full 100-review run: 20 batches of 5, each followed by a 25-second wait, then a final pass that returns all 100 items.

What I found

Evidence from the sheets

These are screenshots of the real summary and spot-check tabs. In the spot-check tab, the theme_ok and urgent_ok columns hold my own Y or N judgments for 33 reviews (every review whose ID divides by 30), and the two counters at the top right add them up.

The summary tab: 100 reviews scored, 100 unique review IDs, 97 sentiment correct (97.00 percent), 41 of 43 positive and 56 of 57 negative reviews identified, 2 neutral answers, 1 flipped, 31 flagged urgent, theme counts totalling 100 and 0 answers outside the three allowed.
The summary tab. The checks that matter: 100 rows, 100 unique IDs, theme counts adding to 100, and no answers outside the allowed values.
The spot-check tab listing sampled reviews with the AI’s sentiment, theme and urgent flag, my Y or N judgments for theme and urgency, a notes column, and counters at the top right reading 24 / 33 and 30 / 33.
The spot-check tab, first rows only. My judgments are in the theme_ok and urgent_ok columns, and the counters read 24 / 33 and 30 / 33.

Limitations

Case study

Support Ticket Triage with n8n and Gemini

An n8n workflow that reads banking support tickets from a Google Sheet, asks Gemini to sort each one into one of 10 categories, and writes the answer back. I scored it against the labels in the public BANKING77 dataset.

Download the full case study (PDF)

Result on 100 tickets

96/100
Matched the dataset label in one scored run
4
Misses, all between neighbouring categories (my reading)
0
Answers outside the 10 allowed categories

What was checked, and how

MeasureResultHow it was checked
Category matches the dataset label96/100Sheet formula comparing Gemini’s answer to the label, yes or no
Answers outside the 10 categories0Every answer was one of the allowed names
Confidence and urgencyNot scoredThese are Gemini’s own labels and the dataset has no answer key for them

How it works

  1. Read 100 tickets from a Google Sheet. They come from the BANKING77 dataset: 10 tickets in each of 10 categories.
  2. Pass only the ticket ID and text on, so the model never sees the dataset’s label.
  3. Send one ticket at a time to Gemini (gemini-3.1-flash-lite, free tier), with a 6-second pause between tickets.
  4. Gemini returns a category, a confidence and an urgency as JSON, and a short script reads it.
  5. Write the answers back to the matching row. A formula marks each row yes or no against the dataset label.
The n8n canvas for the ticket triage workflow after the final run. Message a model, Code in JavaScript, Update row in sheet and Wait each processed 100 items, and Loop Over Items passed on 100 items.
The n8n workflow after the final 100-ticket run.

The four misses

TicketDataset labelGemini said
25: “I am still waiting for a transaction to be completed”transfer_not_received_by_recipientpending_transfer
26: “I just lost my wallet and I see that they are already withdrawing money from my account.”cash_withdrawal_not_recognisedlost_or_stolen_card
55: “How long does it take for a transfer to get to a recipient?”transfer_not_received_by_recipientpending_transfer
95: “how long do money transfers take?”transfer_not_received_by_recipientpending_transfer

What I found

Limitations

Scored results sheet (PDF)  ·  n8n workflow file (JSON, credentials removed)

Automation

Remote Job Radar

A workflow that watches for remote job listings and surfaces only the ones that fit a profile, so the best matches reach me instead of me scrolling through everything.

How it works

  1. Fetch remote job listings from the Remotive API every day.
  2. Score each listing from 0 to 100 with Gemini against my target profile, with a short reason. Gemini also flags whether I can apply from Nigeria.
  3. Filter: keep only matches scoring 65 or higher, and drop roles flagged as closed to Nigeria.
  4. Check each match against the sheet and skip duplicates.
  5. Save new matches to Google Sheets with the score, the reason, the apply link and a status of New.
  6. Send a Telegram alert for each new match.
The Remote Job Radar scenario in Make: an HTTP request, an iterator and text aggregator, a Gemini request, JSON parsing, Google Sheets search and add row steps, then a Telegram message. The scenario is active and runs daily.
The Make scenario, active and scheduled daily. Filters sit before the sheet steps so only good matches and new listings pass.

What it shows

Tested, then tightened

The first version of the filter let weaker roles through, including writing jobs that Gemini’s own reason calls unrelated to my target. I raised the minimum score to 65, cleared the sheet and ran it again. The next run saved a single role.

Remote Job Radar Google Sheet before the change, with six roles scored 50 to 85 including freelance writing roles whose reasons say they are unrelated to AI automation.
Before: scores of 50 to 55 were saved, including roles the “Why it fits” column describes as unrelated.
Remote Job Radar Google Sheet after the change, with one role, Remote Office Assistant, scored 85.
After: with the minimum score at 65, the next run saved one role, scored 85.

Limitations

Automation

Lead Generation Pipeline

An n8n workflow that finds Lagos businesses, scores them as sales leads, saves the best ones to a Google Sheet and sends me a Telegram summary every Monday. It replaces the copy-and-paste work of building a lead list by hand.

Result of one run

37
Businesses with a phone number pulled from OpenStreetMap
35
Hot or Warm leads saved to the sheet
2
National chains filtered out automatically

How it works

  1. Collect: request named Lagos businesses that list a phone number (restaurants, cafes, pharmacies, clinics, schools and similar) from OpenStreetMap.
  2. Standardise: the source stores phone numbers and websites under different tags, so every business is put into the same columns and every number is written as +234…
  3. Score: each business gets 0 to 100 using the rules below, then a tier: Hot (70 and above), Warm (45 to 69) or Cold.
  4. Save: Hot and Warm leads go to Google Sheets. Rows are matched on a unique ID, so running it again updates rows instead of adding duplicates.
  5. Alert: a Telegram message lists the total and the top five leads.
The n8n canvas showing the lead generation workflow, from trigger to Telegram, with every step passing.

How leads are scored

RulePoints
Has a phone number+30
No website (the clearest opening for an automation or web service)+25
Independent business, not a national chain+20
Has an email address+15
Has a street address+15
Appointment-based business (clinic, pharmacy, school and similar)+10
National chain−40
No phone and no emailcapped at 30

The scoring is plain rules, not AI. That keeps it free and fast, and anyone can change a rule and see the effect.

Problems I found and fixed

Limitations

Let’s talk

I am looking for remote roles in automation and AI operations. Download a full case study, or reach me directly.

Receipt extraction case study (PDF) Review intelligence case study (PDF) Ticket triage case study (PDF)

Email: segunmarcus@hotmail.com

LinkedIn: linkedin.com/in/segunmarcus

Built and tested on public data (SROIE receipts, OpenStreetMap business listings, UCI Amazon review sentences). All figures come from my own runs and scoring sheets.