Lead: Avito hiring as a marathon
Avito is more than Russia’s largest classifieds platform. It is also one of the country’s most attractive engineering employers, with hundreds of millions of listings, tens of millions of monthly users, and hard technical problems in search, ranking, fraud detection, and large-scale infrastructure. For developers, an Avito interview can feel like a marathon: structured, deep, and demanding.
This guide breaks down the entire Avito hiring process for 2026, shows what to prepare for at each stage, and explains how to train those stages with mock interviews, quizzes, and task breakdowns on IT Career Gym — not as a slogan, as links in the sections below.
Disclaimer. This is a candidate-side map from public reports and a typical structure. Stages vary by role and grade. IT Career Gym is not an Avito partner and does not promise an offer.
Why engineers want to join Avito
Avito Tech employs more than 3,000 engineers and operates at a scale that few companies can match. The engineering culture is strong, autonomous, and heavily focused on backend systems, search, and machine learning. Scale figures below are a working estimate for prep, not an Avito audit.
| Metric / Area | Typical Scale or Stack |
|---|---|
| Listings on the platform | 100M+ |
| Unique users per month | 50M+ |
| Main backend language | Go |
| Legacy backend | PHP, gradually being phased out |
| Infrastructure | Kubernetes, Kafka, ClickHouse |
| Machine learning | In-house ranking and moderation platforms |
| Engineering culture | Code review, documentation, open technical debt discussions |
| Team structure | Autonomous teams with high ownership |
| Work-life balance | Considered a cultural priority |
What makes Avito different
- Go-first mindset. Most new backend services are written in Go. Deep Go knowledge is not optional.
- Search and ML matter. Ranking, relevance, recommendations, and moderation are central to the product.
- Autonomous teams. Teams make their own technical decisions, which means freedom and responsibility.
- Strong engineering standards. Code review, documentation, and technical debt are treated seriously.
- Healthy work-life balance. Rare for Big Tech, but part of Avito’s culture.
The Avito hiring pipeline at a glance
The full cycle usually takes 3–4 weeks. The exact stages can vary by role and seniority, but the core structure is consistent.
| Stage | Duration | Main Focus | What Success Looks Like |
|---|---|---|---|
| HR screen | 20–30 min | Motivation, experience, expectations | Clear story, product interest, realistic salary range |
| Technical interview | 90 min | Deep stack knowledge | Confident Go, databases, distributed systems, search |
| Algorithms | 60 min | 2 practical tasks | Correct code, clear communication, edge cases |
| System Design | 60 min | Architecture for Senior+ | Requirements, trade-offs, scalability, reliability |
| Final / Team fit | 45–60 min | Culture, soft skills, autonomy | Honest answers, curiosity, strong questions |
Stage 1: HR screen
Format: video call, 20–30 minutes. The HR screen is not just a formality. Avito uses it to understand your motivation, direction, and fit. A clear story is a quiz plus a behavioral mock — not a fake HR bot.
Common HR questions
| Question | What They Want to Hear |
|---|---|
| Tell me about yourself. | A concise, structured career story |
| What do you know about Avito? | Evidence that you researched the product and tech |
| Why Avito? | A specific reason, not “I like big companies” |
| Which direction interests you? | A clear match with your experience |
| What are your salary expectations? | A researched, realistic range |
Avito directions
| Direction | Typical Work |
|---|---|
| Core | Listings, search, user profiles |
| Vertical | Auto, real estate, jobs |
| Monetization | Ads, promotion, paid features |
| Trust & Safety | Anti-fraud, content moderation |
| Platform | Infrastructure, internal tools |
Choose the direction that best matches your background. It increases your chances at every later stage.
Stage 2: Technical interview
Format: video call, about 1.5 hours. This is a deep dive into your stack. Prep the holes in a quiz, then say the answers out loud on Find a partner.
Backend: Go knowledge
| Topic | What to Prepare |
|---|---|
| Concurrency | Goroutines, channels, WaitGroup, errgroup |
| Scheduler | How Go’s M:N threading model works |
| Context | Cancellation, deadlines, propagation |
| Patterns | Worker pool, fan-out/fan-in, pipeline |
| Garbage collection | How GC works in Go and what affects latency |
| Interfaces | Composition, embedding, implicit implementation |
| Testing | Table-driven tests, mocks, benchmarks |
| Profiling | pprof, trace, flame graphs |
Databases
| Database | Key Topics |
|---|---|
| PostgreSQL | Indexes, EXPLAIN ANALYZE, partitioning, MVCC, isolation levels, replication |
| ClickHouse | Columnar storage, analytical workloads, use cases |
| Redis | Data types, persistence, clustering |
Search: critical for Avito
| Topic | What to Understand |
|---|---|
| Full-text search | How a search engine works end to end |
| Inverted index | How terms map to documents |
| TF-IDF / BM25 | Basic ranking intuition |
| Elasticsearch vs PostgreSQL FTS | Trade-offs, scale, relevance |
| Query processing | Tokenization, stemming, typo tolerance |
Distributed systems
| Topic | What to Prepare |
|---|---|
| CAP theorem | Real examples of trade-offs |
| Consensus | Raft, Paxos at a basic level |
| Eventual consistency | How to design around it |
| Distributed transactions | 2PC, Saga |
| Failure handling | Retries, idempotency, timeouts, circuit breakers |
Kafka
| Topic | What to Prepare |
|---|---|
| Consumer groups | How partitions are assigned |
| Offsets | Commit strategies, replay |
| Delivery guarantees | At-least-once, exactly-once |
| Ordering | How to preserve order per key |
| Poison messages | Dead-letter queues, retries, monitoring |
Frontend developer
| Area | Topics |
|---|---|
| JavaScript | Event loop, prototypes, closures, promises, modules |
| React | Fiber, reconciliation, hooks, state management, Server Components |
| Performance | Bundle optimization, tree shaking, code splitting, list virtualization |
| TypeScript | Generics, utility types, discriminated unions, type guards |
Mobile developer
| Platform | Topics |
|---|---|
| Android | Jetpack, Coroutines, memory leaks, Gradle optimization |
| iOS | SwiftUI vs UIKit, Combine, retain cycles, MVVM/VIPER/TCA |
Stage 3: Algorithms
Format: 1 hour, 2 tasks. This is one of the most important sections. Avito tends to prefer Medium-level practical problems over obscure competitive programming puzzles. Communication matters almost as much as correctness. Practice the talk track on Find a partner and the patterns in quizzes.
Core algorithm topics
| Topic | Examples |
|---|---|
| Arrays and strings | Two pointers, sliding window |
| Hash tables | Frequency maps, deduplication |
| Binary search | Search space reduction |
| Trees and graphs | BFS, DFS, traversal |
| Dynamic programming | Basic optimization problems |
| Greedy algorithms | Interval scheduling, local optimum |
Example Avito-style tasks
| Task | What It Tests |
|---|---|
| Find duplicate listings | Hashing, text similarity, threshold logic |
| Best time to buy and sell | Array scanning, profit maximization |
| Category tree grouping | Tree traversal, recursion |
| Listing ranking | Sorting, scoring, decay functions |
How to perform well
- Think out loud.
- Start with brute force, then optimize.
- Discuss edge cases before coding.
- Test with examples.
- Keep your code clean and readable.
Medium tasks die in silence
Two Avito problems in 60 minutes is a communication round as much as a coding round. Rehearse with a partner, then grind the patterns you still fumble.
Stage 4: System design
Format: 1 hour. Usually for Senior+ roles. Rehearse the talk track on Find a partner — the gym does not ship an architecture simulator.
Typical prompts include:
- Search system for listings
- Content moderation service
- Recommendation system
- Anti-fraud system
- User-to-user messenger
- View analytics platform
Example: Avito listing search
Requirements: 100M listings, 10K requests per second, relevance plus freshness, filters for price, city, and category.
| Component | Purpose |
|---|---|
| API Gateway | Rate limiting, authentication |
| Search Service | Query parsing, building search requests |
| Index | Elasticsearch cluster, sharded by city/category |
| Ranker | ML model for relevance, freshness, quality, CTR |
| Cache | Redis for popular queries, short TTL |
| Indexer Pipeline | Kafka-driven near real-time indexing |
What to discuss
- How do you update the index when a listing changes?
- How do you balance relevance and freshness?
- How do you handle typos in queries?
- How does the system scale as traffic grows?
- What are the failure modes and fallbacks?
Stage 5: Final interview
The final stage is about team fit, motivation, and how you work with others. Bank STAR stories and run them on a pair mock.
| Question | What They Evaluate |
|---|---|
| Tell me about a difficult technical decision. | Judgment, ownership, trade-offs |
| How do you work with uncertainty? | Adaptability, planning |
| How do you give and receive feedback? | Communication, maturity |
| What matters to you in a team? | Culture fit, collaboration |
Questions you should ask
- What are the biggest technical challenges in the team?
- How are processes organized?
- What does onboarding look like?
- How does growth work here?
Common candidate mistakes
| Mistake | How to Avoid It |
|---|---|
| Not knowing the product | Use Avito as a user before interviewing |
| Weak Go knowledge | Go is the main language; prepare deeply |
| Silence during algorithms | Think out loud and explain trade-offs |
| Not asking questions in System Design | Ask clarifying questions early |
| Underestimating the final interview | Prepare behavioral stories |
2-week preparation plan
Week 1: Theory
| Day | Topic |
|---|---|
| 1–2 | Go: concurrency, patterns, testing |
| 3–4 | PostgreSQL, Redis, ClickHouse |
| 5–6 | Distributed systems, Kafka |
| 7 | Elasticsearch and search fundamentals |
Week 2: Practice
| Day | Topic |
|---|---|
| 1–3 | LeetCode: 20 Medium tasks |
| 4–5 | System Design: search, recommendations |
| 6 | Mock interview — or a challenge if you need a deadline |
| 7 | Study Avito products and user flows |
How IT Career Gym maps to stages
The gym does not book an Avito interviewer. It matches you with another engineer and gives you quizzes and artifacts you can bring to that call.
| Avito Stage | IT Career Gym |
|---|---|
| HR screen | Salary and story quizzes, then a behavioral mock |
| Technical interview | Deep stack quizzes and a 90-minute pair mock |
| Algorithms | Partner practice on Find a partner; bring a pet-project if you want a shared artifact |
| System Design | Architecture talk track on a pair mock — listing search, not a simulator |
| Final interview | STAR stories plus a culture-fit mock |
Final thoughts
The Avito interview is a marathon, not a sprint. The company values deep technical knowledge, clear communication, and cultural fit. Start preparing early, focus on Go, databases, algorithms, and system design, and practice explaining your thinking out loud.
With the right preparation — and a real mock, not a second reading of this page — you can move through every stage with a clearer talk track.
Questions people actually ask
Tap a row. Only one stays open — these are stage and prep answers, not a glossary of what Avito is.
-
Do I need Go if I interview as PHP, frontend, or mobile?
Go is required for new backend services. Frontend and mobile loops go deep on their own stacks. PHP is being phased out — name it as legacy context, not as the language you plan to write.
Link to this answer -
How long is the full Avito cycle, and do all five stages always happen?
The loop usually takes 3–4 weeks. HR, a 90-minute stack interview, and algorithms are the core. System design is typically Senior+. The final culture round is common, but teams skip or merge stages.
Link to this answer -
Is this Hard LeetCode or Medium plus communication?
Medium practical tasks, two in an hour. Think out loud, start brute force, then optimize. Silence fails as often as a wrong answer.
Link to this answer -
Is System Design only for Senior+?
Usually yes: one hour on listing search, moderation, recommendations, anti-fraud, messenger, or view analytics. Middle grades should still sketch requirements and trade-offs.
Link to this answer -
Should I name a specific vertical on the HR screen?
Yes if it matches your background: Core, Auto, real estate, jobs, monetization, Trust & Safety, or Platform. A vague “I like classifieds” is weaker than a direction you can defend.
Link to this answer -
Does IT Career Gym book a slot with an Avito interviewer?
No. The gym is not an Avito partner and does not promise an offer. Matching finds another engineer for a pair mock. You run the call yourselves.
Link to this answer -
What should I do in two weeks if I already have an interview date?
Week 1: Go, databases, Kafka, search. Week 2: Medium tasks, one listing-search design, a mock on day 6, then Avito product flows. Use a deadline challenge if you need pressure.
Link to this answer
Train the loop, not the slogan
Open a search, run a Go or listing-search mock, and keep the quizzes for the holes that still stall you. Already have an account? Log in and start matching.
Also in this series: Pair mock interview · Last week before the interview · Amazon system design · Stripe system design · STAR stories