Pair-Programming AI Booking MVP Leads
Budget / Salary€750–1,500
TypeFreelance project
LocationRemote
Posted45 minutes ago
Bids are valid only through our test. Bids without a completed test and its receipt number will not be considered.
Test: https://artusihabibot.com/dev-assessment/
Access passphrase: Nazimerd
About the project
Artusi Ħabibot is a multilingual, chat-first restaurant booking assistant: users find a place, request a table and receive a real confirmation and a reminder. We start on Telegram and will extend to WhatsApp and voice calls. No app download. The goal is a working MVP in a pilot city in Italy.
Who we are looking for
Two lead engineers who will work together in pair programming and practise test-driven development (failing test first, then code, then refactor). Tests are how we agree scope and acceptance, so management stays light. No one approves their own work: each lead reviews the other’s evidence.
Main stack: Python (aiogram, async), PostgreSQL, LLMs via OpenRouter or open models, telephony and speech-to-text/text-to-speech for voice.
Phases and budget (per person)
Phase 1 – Plan validation: USD 50. Complete the plan together and confirm the development scope, the infrastructure, telephony and AI costs, and a realistic marketing and user-acquisition hypothesis.
Phase 2 – Build: USD 200, starting in December. Only if the Phase 1 plan is validated.
Phase 3 – Pilot: USD 200, starting in January.
Fixed payments go through Freelancer.com milestones, released for accepted and tested deliverables. Beyond the fixed budget, compensation is a revenue fee per confirmed booking with a cap, or shares for those who want to stay in the project long term.
How to apply
Open the test, enter the passphrase above and your email, and type the 6-digit code you receive.
Answer four technical challenges and a cost plan for calls and AI (sourced unit prices, monthly totals, cost per booking), plus your motivation, proposed economics and an AI-use declaration. Drafts are saved automatically.
Submit once and quote your receipt number (HB-…) in your bid, together with your weekly availability for pairing and your time zone.
The evaluation criteria and weights are visible inside the test before you answer.
Important: the name on your Freelancer.com profile and the name on the email account you use for the test must match. Mismatching identities will be reported to Freelancer.com.
Test: https://artusihabibot.com/dev-assessment/
Access passphrase: Nazimerd
About the project
Artusi Ħabibot is a multilingual, chat-first restaurant booking assistant: users find a place, request a table and receive a real confirmation and a reminder. We start on Telegram and will extend to WhatsApp and voice calls. No app download. The goal is a working MVP in a pilot city in Italy.
Who we are looking for
Two lead engineers who will work together in pair programming and practise test-driven development (failing test first, then code, then refactor). Tests are how we agree scope and acceptance, so management stays light. No one approves their own work: each lead reviews the other’s evidence.
Main stack: Python (aiogram, async), PostgreSQL, LLMs via OpenRouter or open models, telephony and speech-to-text/text-to-speech for voice.
Phases and budget (per person)
Phase 1 – Plan validation: USD 50. Complete the plan together and confirm the development scope, the infrastructure, telephony and AI costs, and a realistic marketing and user-acquisition hypothesis.
Phase 2 – Build: USD 200, starting in December. Only if the Phase 1 plan is validated.
Phase 3 – Pilot: USD 200, starting in January.
Fixed payments go through Freelancer.com milestones, released for accepted and tested deliverables. Beyond the fixed budget, compensation is a revenue fee per confirmed booking with a cap, or shares for those who want to stay in the project long term.
How to apply
Open the test, enter the passphrase above and your email, and type the 6-digit code you receive.
Answer four technical challenges and a cost plan for calls and AI (sourced unit prices, monthly totals, cost per booking), plus your motivation, proposed economics and an AI-use declaration. Drafts are saved automatically.
Submit once and quote your receipt number (HB-…) in your bid, together with your weekly availability for pairing and your time zone.
The evaluation criteria and weights are visible inside the test before you answer.
Important: the name on your Freelancer.com profile and the name on the email account you use for the test must match. Mismatching identities will be reported to Freelancer.com.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.