Azizi's product in action
Local-language voice IVR that compresses weeks of cross-departmental work into a five-minute phone call — two demos.

Azizi's product in action
2G USSD and IVR interactions remain prevalent in the global south, especially Africa, where over 40% of mobile connections in Sub-Saharan Africa are still on 2G networks. This supports high volumes of basic feature-phone usage for services like mobile money and airtime top-ups, contrasting sharply with smartphone-dominated interactions elsewhere. In South Africa, for instance, smartphone users spend up to 10% of their time on 2G/3G, while 14 million customers and 6 million annual 2G device sales highlight persistent reliance. [Source — OpenSignal]
What are we doing with that?
We help companies quickly onboard AI-powered IVR (interactive phone calls), SMS and USSD experiences that adapt in real time, with or without systems-integration readiness. Why? To reach the underserved employees and users that would be "too expensive" to build for.
Here is a taste of our capabilities.
Use Case 1: Log a repair job card
HelloTractor use case. We created this trial of a simulated phone call for a job-card scenario — no app required, just call a number.
Phone call (no internet required)
Example log (Google Sheets integration)

Conversation (with auto-translation)
Assistant. Karibu Hello Tractor! Are you an Operator, Technician, Farmer, or Booking Agent? — Welcome to Hello Tractor! Are you an Operator, Technician, Farmer, or Booking Agent?
Caller. Mini ni technician. Leo nimefanya repair kwa tractor KBZ 412. Nilibadilisha oil filter, part number AF-2035, nikaweka engine oil 10 liters, na nikakaza fan belt. — I am a technician. Today I repaired the tractor KBZ 412. I replaced the oil filter, part number AF-2035, added 10 liters of engine oil, and tightened the fan belt.
Assistant. Nimekuelewa, umefanya repair kwa tractor KBZ 412, ukibadilisha oil filter part number AF-2035, kuweka engine oil 10 liters, na kukaza fan belt. Sasa, nipe masaa mangapi umetumia kwa kazi hii? — I understand you have repaired the tractor KBZ 412 by changing the oil filter part number AF-2035, adding 10 liters of engine oil, and tightening the fan belt. Now, how many hours did you spend on this work?
Technician. Mimi nilitumia masaa mawili kumaliza kazi hii. — It took two hours to finish this work.
Assistant. Asante. Umetumia masaa mawili kwa kazi hii. Nimejaza job card: repair on tractor KBZ 412. Work: oil filter change, engine oil addition, fan belt tightening. Parts: AF-2035. Sawa? Nitasave sasa. — Thank you. You have spent two hours on this work. I have filled out the job card: repair on tractor KBZ 412. Work: oil filter change, engine oil addition, fan belt tightening. Parts: AF-2035. Is that okay? I will save the details now.
Technician. Ndio. Wazi, asante sana. Kwaheri. — Yes. Clear, thank you very much. Goodbye.
What the agent did behind the scenes
| Turn | Action |
|---|---|
| 3 | Extracted structured data and logged the repair details |
| 3 | Wrote a row to a Google DB |
| 3 | Sent a WhatsApp summary |
Extracted and formatted for API integration:
{
"Phone": "254722***440",
"Timestamp": "2023-10-12T12:00:00Z",
"Tractor_ID": "KBZ 412",
"Location": "",
"Job_Type": "repair",
"Work_Description": "Oil filter change, engine oil addition, fan belt tightening",
"Parts_Used": "AF-2035",
"Labor_Hours": "2",
"Missing_Info_Flag": "FALSE",
"Status": "Completed"
}A WhatsApp message is sent to the technician:
An alert can also be sent to other stakeholders.
Use Case 2: Small entrepreneur insights
For a non-profit that's built a tool to help businesses track their inputs and outputs, we are inferring and guiding on how to use the tool.

Phone call (no internet required)
Conversation (with translations)
Advisor. Habari. Mimi ni Digikua kutoka Somo Buruka. Niko hapa kukusaidia na biashara yako. Tuanze wapi leo? — Hello. I am Digikua from Somo Buruka. I am here to help you with your business. Where should we start today?
Alfred. Habari Digikua. Mimi ni Alfred kutoka Ethan Bakery, Northern Tanzania. Naomba insights za biashara yangu. — Hello Digikua. I am Alfred from Ethan Bakery, Northern Tanzania. I would like insights for my business.
Advisor. Habari Alfred. Ngoja nichukue data yako. — Hello Alfred. Wait, let me get your data…
Alfred. Sawa. — Okay.
Advisor. Nimeangalia data ya Ethan Bakery. Revenue yako ni about three hundred fifty two thousand Shillings, na margin ni ninety nine point six percent. Una sales five hundred ninety eight. Unataka tuangalie nini zaidi? — I have checked the data for Ethan Bakery. Your revenue is about three hundred fifty-two thousand Shillings, and the margin is ninety-nine point six percent. You have sales of five hundred ninety-eight. What else do you want us to look at?
Alfred. Ningependa kuelewa revenue trend kwanza — nimeona tumeshuka mwaka huu. — I would like to understand the revenue trend first — I have seen that we have decreased this year.
Advisor. Revenue yako imeshuka fifty one percent year over year. Hii inaonyesha kuna changamoto. Jaribu kuangalia kama wateja wako wanarudi ama kuna customer churn. Module 6 ya Customer & Market Focus inaweza kusaidia. Unataka kuendelea na kitu kingine? — Your revenue has decreased by fifty-one percent year over year. This indicates that there are challenges. Try to check if your customers are returning or if there is customer churn. Module 6 of Customer & Market Focus may help. Do you want to continue with something else?
Alfred. Na mimi ninafanya vipi ikilinganishwa na wengine kwenye Somo programme? — How do I compare to others in the Somo programme?
Advisor. Nimeangalia peers wako katika Somo programme. Samabo Mila ana margin ya eighty five point nine percent, na Neighbourhood Bakery iko at thirty nine point five percent. Hii ni more typical kwa bakeries. Uko sawa na hiyo? — I have checked your peers in the Somo programme. Samabo Mila has a margin of eighty-five point nine percent, and Neighbourhood Bakery is at thirty-nine point five percent. This is more typical for bakeries. Are you okay with that?
Alfred. Asante. Labda sijaingiza gharama zangu zote kama ingredients na rent. Nitaupdate records zangu leo. Ninahitaji msaada gani kuingia DigiKua? — Thank you. Maybe I haven't included all my expenses like ingredients and rent. I will update my records today. What help do I need to access DigiKua?
Advisor. Nimeona margin yako ya ninety nine point six percent ni high sana kwa bakery. Most bakeries zinakuwa at thirty to fifty percent. Labda haujaingiza gharama zako zote kwa DigiKua. Kuhusu ku-update records zako, unaweza kuingia kwenye DigiKua platform, nenda kwa sehemu ya expenses, na ongeza gharama kama ingredients, rent, na electricity. Umefika hapo? — I have seen that your margin of ninety-nine point six percent is very high for a bakery. Most bakeries are at thirty to fifty percent. Maybe you haven't entered all your costs into DigiKua. Regarding updating your records, you can log into the DigiKua platform, go to the expenses section, and add costs like ingredients, rent, and electricity. Have you reached that point?
Alfred. Ndio, nimefahamu kila hatua. — Yes, I have understood every step.
Advisor. Asante Alfred. Nimeweka insights zako na data quality flag kwa records. Utapokea summary ya mazungumzo yetu kwenye WhatsApp. Kuna kitu kingine unahitaji? — Thank you, Alfred. I have added your insights and the data-quality flag to the records. You will receive a summary of our conversation on WhatsApp. Is there anything else you need?
Alfred. Asante sana Digikua. Kwaheri. — Thank you very much, Digikua. Goodbye.
What the bot did behind the scenes
| Turn | Action |
|---|---|
| 2, 3 & 5 | Pulled user data |
| 7 | Wrote a log |
| 7 | Sent a WhatsApp summary |
What API data is generated (at turn 7) — raw row data, could be ingested straight into the system:
{
"Session_Date": "2023-10-10",
"Business_Name": "Ethan Bakery",
"Owner_Name": "Alfred",
"Category": "Food and Beverage",
"Region": "Northern Tanzania",
"Country": "Tanzania",
"Insight_Category": "Revenue_Trend",
"Key_Insight": "Revenue imeshuka fifty one percent year over year. Investigate customer churn.",
"Recommended_Action": "Module 6 (Customer & Market Focus)",
"Data_Quality_Flag": "TRUE",
"Session_Outcome": "Completed"
}Log and API: Data_Quality_Flags (saved at turn 7):
{
"Flag_Date": "2023-10-10",
"Business_Name": "Ethan Bakery",
"Flag_Type": "implausible_margin",
"Field_Affected": "Margin_Pct",
"Suggested_Fix": "Add missing expenses like ingredients, rent.",
"Learner_Committed_To_Update": "TRUE",
"Resolution_Status": "Pending"
}WhatsApp message sent:
![A WhatsApp message reading: [DEMO] Digikua session summary for 2023-10-14. Topic: Revenue Trend. Action: Investigate customer churn. Next: Update expenses on DigiKua.](/articles/azizi/product-in-action/whatsapp-summary.png)
How to join in making history
If you got here, we don't just appreciate it — you can get involved, and earn. If you're more interested in how we got here, and how we make this work, read our culture-as-a-product article.
Join us in building AI abundance for those who care about the global south, today.
Come build with us — hello@aziziafrica.com