Viktor Nirov
LLM Developer
8+ years exp
Lua
C#
Python
Active 16 days ago
+7 (967) 5417338 Invite to interview
Message
Download CVCV
Overview
Technical skills
Timeline
Roles
Overview
A capable developer who produces clean, self-contained tooling and educational/game code at a middle (early mid-level) competence. The strongest proven skill is Python CLI/tooling and text-processing, demonstrated by the analyze_lessons.py script (argparse-driven CLI, modular analysis, CSV output and reporting). There is limited evidence of backend system design - no databases, network APIs, scalability patterns or production-grade reliability/observability features are present in the analyzed human-authored files.
Technical skills
Lua
C#• Senior • 8y+
Python• Middle
C#
.NET• 8y+
Python
Requests
Beautiful Soup
AI/ML
Claude
Embeddings
Gemini
Qwen
AI Agents
LLM
Prompt Engineering
PyTorch
Transformers
Mobile
MVP
MVVM
DevOps
CI/CD
Rest API• 8y+
Git
Game Dev
ShaderLab
Unity• 8y+
Timeline
Software Engineer (AI / Agent Systems)
•
Middle
NDA
•
Full-Time
Designed end-to-end AI agent workflows for a commercial product under NDA, including agent roles, behavior, orchestration and context control. Integrated tool/function calling with Skills and MCP-style tooling, configuring memory, retries, fallbacks and guardrails. Implemented API-based integration with product systems using structured contracts, handling errors to ensure predictable action execution. Took part in the full engineering cycle from hypothesis and prototyping to testing, rollout and support.
AI Agents
LLM
Prompt Engineering
Rest API
Unity Developer / C# Developer
•
Middle
SOFTINTERMOB
•
Full-Time
Built Unity/C# mobile game features from prototype to release, implementing gameplay logic, UI/client logic, runtime systems and integration layers. Managed live production support by investigating regressions and performance/UI/gameplay bugs and delivering fixes to verifiable outcomes. Designed client architecture for new mechanics by separating gameplay, UI, services, configuration, networking and integration layers. Integrated external assets and services such as Ads, Analytics, IAP and REST/API integrations, and improved CPU/GPU/RAM usage, scene loading, build size and responsiveness using profiling.
Unity
C#
MVVM
Rest API
Unity Developer / Senior C# Developer
•
Senior
NDA
•
Full-Time
Worked across multiple commercial mobile/NDA Unity/C# projects, delivering gameplay mechanics, core client systems, UI flows and data-driven configuration. Implemented VR and multiplayer gameplay behaviors including session flow, party/invite logic, state synchronization and reconnect/reset edge cases. Built reusable Unity modules for inventory/shop/economy, timers, UI animation, pooling, save/config flows, analytics events and SDK wrappers to speed up feature delivery. Integrated external APIs and runtime services, refactored legacy code for better boundaries and testability, and handled production support with profiling and root-cause analysis to stabilize performance and gameplay behavior.
Unitysince 2018
C#since 2018
Rest APIsince 2018
Ural State University of Economics
Bachelor's Degree •
Business Informatics
Middle Backend Developer
Confidence: Medium Generalist
A capable developer who produces clean, self-contained tooling and educational/game code at a middle (early mid-level) competence. The strongest proven skill is Python CLI/tooling and text-processing, demonstrated by the analyze_lessons.py script (argparse-driven CLI, modular analysis, CSV output and reporting). There is limited evidence of backend system design - no databases, network APIs, scalability patterns or production-grade reliability/observability features are present in the analyzed human-authored files.
API Design
2/10
How well APIs are designed
Minimal API/CLI design present - a well-structured CLI script (argparse) exists but there is no networked API, versioning or error contract design.
Data Layer & Database
1/10
Working with databases
No database layer or migrations - only file/CSV data handling for lesson artifacts.
Scalability & Performance
1/10
Handling load and speed
No real scalability or measured performance engineering for backend systems; only small-scale text processing and limits in scanning logic.
System Architecture
2/10
Overall system structure
Simple modular tooling with clear function boundaries (good for a CLI utility) but no multi-service decomposition, config/secrets management, or runtime fault-isolation design.
Security & Auth
1/10
Protecting data and access
Almost no security-specific engineering for backend concerns - only basic input/path checks when locating lesson files.
Reliability & Observability
2/10
Stability and monitoring
Basic reliability patterns in a CLI tool - return codes and some error checking; no structured logging, metrics, retries or graceful shutdown for services.
Expertise
Python• Middle
.NET• Junior
Industries
Education• Middle
Gaming• Middle
Recommendations
- Build CLI and developer tooling for content pipelines (lesson validation, linting, CSV/JSON export) where their analyze_lessons.py expertise directly applies.
- Lead small-to-medium Python projects that require parsing, validation and report generation (content QA tooling, static analysis utilities, or course-build assets).
- Develop and extend game-focused components or education tooling in C#/.NET (Unity lesson scripts, test harnesses, automation for course verification), starting with incremental automation and test coverage.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories:
Junior AI/ML Engineer
Confidence: Medium Generalist
A practical generalist software engineer at an early-to-mid career level with solid Python tooling and web-integration experience. The strongest proven skill is reliable web data ingestion and tool integration, demonstrated by fetch_url_text and the FastMCP tool wiring in web-search-neo/main.py along with the search/parsing implementations in web-search-neo/msp_search.py. There is no public evidence in the human-authored files of ML model training, custom architectures, advanced MLOps, or GPU/quantization work.
Model Architecture & Training
How well models are designed and trained
Not evidenced in public code
Data Pipeline & Feature Engineering
2/10
How data is prepared for models
Implements web data ingestion and parsing (HTML fetching, text extraction, search-result scraping) with error handling and timeouts.
Evidence
web-search-neo/main.py:fetch_url_text
web-search-neo/msp_search.py:search_google
Experimentation & Evaluation
2/10
How results are measured and tested
Basic experimentation/evaluation artifacts - unit tests exercising the fetch/search tools and guards for flaky external dependencies.
Evidence
web-search-neo/test_msp_server.py
MLOps & Deployment
2/10
How models are shipped to production
Service wiring and deployment-ready tooling visible - a FastMCP tool server exposing fetch/search endpoints and a clear main() entrypoint.
Evidence
web-search-neo/main.py:@mcp.tool() definitions and main()
web-search-neo/msp_search.py:get_driver (selenium driver bootstrapping)
Computational Efficiency
How efficiently computing resources are used
Not evidenced in public code
Research Depth & Innovation
Depth of research and new ideas
Not evidenced in public code
Expertise
RAG• Junior
Industries
Internet Services• Junior
Technologies
Python• Middle
Rest API• 8y+
Claude
Qwen
Embeddings
Prompt Engineering
Beautiful Soup
AI Agents
CI/CD
Transformers
Git
PyTorch
Gemini
LLM
Requests
Claude• mentioned only
Claude Code• mentioned only
Gemini• mentioned only
Recommendations
- Build backend tools that ingest and normalize web text for retrieval-augmented pipelines (implement connectors, deduplication, and canonicalization).
- Develop small web-service features and microtools: extend FastMCP tools, add robust retries/backoff, timeouts, and monitoring for scraping jobs.
- Work on integrating these ingestion tools into lightweight RAG flows or downstream text pipelines (indexing, basic metadata extraction, safe sanitization).
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories:
Junior Game Developer
Confidence: Medium Gameplay Engineer
A junior-level gameplay engineer who produces clean, lesson-focused Unity MonoBehaviour scripts for small game mechanics. The strongest proven skill is implementing gameplay logic and UI wiring, evidenced by TicTacToeGame (multiple lesson versions) and ClickerGame (click feedback, audio, UI). Advanced areas are not evidenced - there is no custom rendering/shader development, no deterministic simulation or measured performance work, and no editor tooling shown in the analyzed human-authored files.
Gameplay Systems & Mechanics
3/10
How game logic works
Clear evidence of small gameplay systems and lesson-level mechanics (turn-based TicTacToe, clicker game) implemented as readable MonoBehaviour scripts, but no advanced architecture or testing.
Graphics & Rendering
1/10
Drawing game visuals
No original rendering or shader work authored; only scene/layout assets and third-party shaders are present.
Physics & Math
1/10
Game physics and math
Basic use of array/list logic and game-state checks (turn switching, win detection) but no custom physics, spatial partitioning or deterministic simulation.
Engine Proficiency
2/10
Skill with the game engine
Shows practical Unity engine usage (MonoBehaviour lifecycle, IPointerDownHandler, AudioSource usage) appropriate for small projects and course material, but no custom editor tooling or advanced package design in the analyzed human-authored files.
Performance & Frame Budget
1/10
Keeping the game smooth
No measured/performance-driven work; contains hot-path allocations (Instantiate/Destroy in click feedback) and no profiler/zero-alloc efforts.
Content Pipeline & Tooling
1/10
Tools for game content
Repository shows organized lesson scripts and course structure for Unity lessons, suitable for content delivery but lacking automated importers, CI or complex pipeline tooling in the scoped human-authored files.
Verified artifacts
Expertise
Gameplay & Mechanics Development• Junior
Industries
Gaming• Middle
Technologies
Lua
Unity• 8y+
ShaderLab
Unity• mentioned only
Unity• mentioned only
Recommendations
- Assign to implement and iterate single-player gameplay features and small game systems (e.g., UI-driven interactions, simple AI for turn-based games).
- Develop course or tutorial content for Unity (lesson scripts, step-by-step examples, and student-facing exercises) leveraging the existing lesson structure.
- Improve existing prototypes by replacing Instantiate/Destroy hot paths with pooling and add simple profiler checkpoints to demonstrate measurable performance gains.
- Extend the TicTacToe/Clicker examples with unit tests and small editor utilities (e.g., playtest helpers, auto-assign scripts) to raise engine-proficiency evidence.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories: