Nativa Global - Challenge

Job Application Challenge
Company: Nativa Global
Role: Bot Developer Manager
Date: June 2024
Description: Four exercises on feasibility, integrations and attention to an incident. Analyze each case and prepare an explanation and if needed a workaround.
Exercise 1
A customer in the retail vertical requires a WhatsApp bot that functions as a lead bot, the bot should ask the user's name, telephone number, email address and location of the development/subdivision of interest.
Additionally, as a mandatory requirement you need to integrate with HubSpot to store these leads.
The client requests the viability of this last point ASAP.
- Is it possible to integrate the bot with hubspot?
- What does Nativa Global need to do it?
- What does the client need?
- If it is not possible to integrate, what workaround could we offer?

Exercise 2
A customer in the financial vertical needs to integrate its services into the bot, such as, for example, identifying clients by phone number, downloading account statements, movements, etc.
Being a company in the financial vertical, according to the integration policies with external services, the bot must follow the highest security standards, which is why they are requesting that communication be through VPN if possible.
On the part of the commercial team, two platform proposals have been offered to the client to develop the bot: Infobip and Liveperson, so it is necessary to be certain of the viability of the project on each platform.
- Infobip official documentation here.
- Official Liveperson Documentation here.
Answer the following questions:
- Is it feasible to meet the VPN communication requirement with the client's APIs? Just because? Why not?
- If yes, what is needed?
- If not, what alternatives do we offer? Take into account that the client requires high security standards.
- Which of the platforms do you recommend to carry out the project?

Exercise 3
You have a new IA project where the bot should use different documents for different products, they are having the issue that the doc is very similar for all their products so when a customer ask something the bot answer randomly with the firs doc that it found.
Please provide the best practices to have a good experiences with the bot and what you will consider for a IA project (Google IA) and a possible conclusion about what it is happening.



Exercise 4
The client reports errors in user conversations.
The bot has 2 API integrations:
- to consult personal data
- to modify said data
For this exercise the APIs point to mock services that simulate the client's real APIs.
The bot went into production weeks ago and the only change recently made by our development team was a text adjustment to the initial greeting.
The client is very upset since all his queries are affected, he mentions that the bot is useless and is considering stopping future related projects until the issue is resolved and assures that the problem originated from the bot's side.
We urgently require that you do a global analysis of the problem, obtaining root cause and conclusions. For the next session with the client, it is required to have this analysis ready, with conclusions, possible solutions or alternatives either from the client's side or from Nativa Global.
Consider the tense situation with the client in your explanation and seek to explain the cause and solutions in the best possible way.
It is recommended that you carry out all the necessary tests, with all the tools that you consider, in order to be certain of the root cause, as well as to have all possible evidence ready (Audit, logs, screenshots, diagrams, documentation, etc.).



