Tochka¶
English · Русский · AI Interview Atlas › Companies
Fintech and payments · Russia · Loop reviewed 2026-09-29
The bank's prompt-engineering team describes how it hires prompt engineers: a questionnaire, an HR call, a week-long prompt-writing test, a technical interview and an interview with the internal client. In 2024, more than 96% of candidates first heard of function calling at the interview.
On this page: Interview loop · Official preparation material · Questions (2)
Roles covered¶
- AI Engineering: Applied AI engineer
Interview loop¶
| Stage | What happens | Basis |
|---|---|---|
| Questionnaire and HR call | A questionnaire, then an HR call about experience with simple questions about OpenAI tools. | ✅ confirmed by the company Бесполезные курсы и помешательство на GPTs: как мы искали prompt-инженеров, Точка, блог на Хабре, published 2024-10-31, retrieved 2026-09-29 |
| Test assignment | A week to write prompts for each type of task the team handles; one example is extracting a date range from a request in Russian, where near-identical phrasings mean different periods. | ✅ confirmed by the company Бесполезные курсы и помешательство на GPTs: как мы искали prompt-инженеров, Точка, блог на Хабре, published 2024-10-31, retrieved 2026-09-29 Как нанимаем, грейдируем и решаем типовые задачи prompt-инженеров в Точке, Точка, блог на Хабре, published 2025-06-26, retrieved 2026-09-29 |
| Technical interview | With the lead of the prompt engineers, HR, a data scientist and a team lead: technical questions and a conversation about experience. | ✅ confirmed by the company Бесполезные курсы и помешательство на GPTs: как мы искали prompt-инженеров, Точка, блог на Хабре, published 2024-10-31, retrieved 2026-09-29 |
| Client interview | An interview with the internal client. In 2024, 23 of 219 candidates reached it and 15 received and accepted offers. | ✅ confirmed by the company Как нанимаем, грейдируем и решаем типовые задачи prompt-инженеров в Точке, Точка, блог на Хабре, published 2025-06-26, retrieved 2026-09-29 |
Coding: unknown
Official preparation material¶
- Бесполезные курсы и помешательство на GPTs: как мы искали prompt-инженеров, Точка, блог на Хабре, published 2024-10-31, retrieved 2026-09-29
- Как нанимаем, грейдируем и решаем типовые задачи prompt-инженеров в Точке, Точка, блог на Хабре, published 2025-06-26, retrieved 2026-09-29
Questions reported for Tochka¶
Agents, tools and protocols¶
- What are a language model's generation parameters?
- Knowledge · Asked at: Tochka ✅
- Tests: Whether you distinguish model generation settings from instructions about role, style or response structure.
- A strong answer covers:
- Name the decoding settings: temperature, top-p, top-k, maximum output tokens and stop sequences.
- Separate them from prompt instructions on role, tone and format; an enforced output schema is an API setting.
- Show each setting's effect on repeated samples, and check which ones a given API actually accepts.
- Read: The Curious Case of Neural Text Degeneration (Holtzman et al., arXiv)
Applied and customer scenarios¶
- Write a prompt that extracts a date range from a customer's request in Russian and tells apart near-identical phrasings such as «с первого числа» (from the first), «по первое число» (through the first) and «за первое число» (for the first).
- Applied scenario · Asked at: Tochka ✅
- Tests: Whether you handle language-specific ambiguity and test a prompt on hard cases.
- A strong answer covers:
- Define the output: start and end dates, inclusive or exclusive bounds, and the reference date.
- Spell out how the prepositions «с», «по» and «за» set each bound, with contrasting examples.
- Test on a set of near-identical phrasings and relative dates, and ask back when it is ambiguous.
- Read: Evaluation best practices (OpenAI API documentation)