Basic requirements (education, skills)

• Experience of fine-tuning and/or RAG systems based on open-source LLM (Qwen, Mistral, LLaMA, Yi or analogues)
• Knowledge of frameworks: vLLM / TGI, LangChain / LlamaIndex, HuggingFace Transformers
• On-premise GPU deployment experience 
• Strong command of Python; experience writing REST API integrations
• Understanding the principles of working with personal data: anonymization, encryption (TLS 1.3, AES-256)
• Experience in developing classifiers and sentiment analysis models based on Russian-language texts

The advantage will be:

• Experience with voice stacks: ASR/TTS (SaluteSpeech, Vosk, Silero) + SIP/VoIP integration
• Knowledge of the requirements of FZ-152 (personal data) and FZ-230 (collection)
• LLM security system design experience (guardrails, prompt injection defense)
• Working with JSON format data + dialog metadata
• Experience in A/B testing of ML models in production

What you have to do

It is necessary to implement the development of a custom LLM system with hyperpersonalization function to automate communications with debtors. The system should be fully deployed in the on-premise infrastructure of the customer. 

Tasks:

• Selection and configuration of the basic LLM for the tasks of dialog recovery (context from 32k to 128k tokens, Russian language only)
• Fine-tuning models based on 50,000+ marked-up dialogues (categories: payment/ restructuring/rejection, sentiment, critical scenarios)
• Development of a knowledge-based RAG pipeline (regulations) for question-and-answer mode
• Implementation of the intent classifier and the dialogue history summarization module (50+ sessions per debtor)
• Development of a three-level generation security system: system prompt → online validator → dialog audit
• Protection against prompt injection through input data sanitization
• Integration via the REST API with the collection system (obtaining debt data, synchronizing the time zone to comply with time windows)
• Ensuring performance requirements: <1.5 seconds per text, up to 200 simultaneous dialogues​
• Participation in the design of the voice module (ASR/TTS + SIP, latency ≤2.5 seconds) 
• Development of quality metrics: LSS ≥ 99.5%, the proportion of escalations to the operator < 30%, uplift recovery rate ≥ 8%

What we offer (Social package, benefits)

• Format: project employment, remotely
• Duration: Stage 1 — 2026 (pilot for 10,000 debtors), further possible continuation
• Infrastructure: on-premise GPU cluster on customer's equipment (Dell PowerEdge), access to a dataset of 200,000+ records and 50,000+ dialogues

Basic requirements (education, skills)

• Experience of fine-tuning and/or RAG systems based on open-source LLM (Qwen, Mistral, LLaMA, Yi or analogues)
• Knowledge of frameworks: vLLM / TGI, LangChain / LlamaIndex, HuggingFace Transformers
• On-premise GPU deployment experience 
• Strong command of Python; experience writing REST API integrations
• Understanding the principles of working with personal data: anonymization, encryption (TLS 1.3, AES-256)
• Experience in developing classifiers and sentiment analysis models based on Russian-language texts

The advantage will be:

• Experience with voice stacks: ASR/TTS (SaluteSpeech, Vosk, Silero) + SIP/VoIP integration
• Knowledge of the requirements of FZ-152 (personal data) and FZ-230 (collection)
• LLM security system design experience (guardrails, prompt injection defense)
• Working with JSON format data + dialog metadata
• Experience in A/B testing of ML models in production

What you have to do

It is necessary to implement the development of a custom LLM system with hyperpersonalization function to automate communications with debtors. The system must be fully deployed in the on-premise infrastructure of the customer. 

Tasks:

• Selection and configuration of the basic LLM for the tasks of dialog recovery (context from 32k to 128k tokens, Russian language only)
• Fine-tuning models based on 50,000+ marked-up dialogues (categories: payment/ restructuring/rejection, sentiment, critical scenarios)
• Development of a knowledge-based RAG pipeline (regulations) for question-and-answer mode
• Implementation of the intent classifier and the dialogue history summarization module (50+ sessions per debtor)
• Development of a three-level generation security system: system prompt → online validator → dialog audit
• Protection against prompt injection through input data sanitization
• Integration via the REST API with the collection system (obtaining debt data, synchronizing the time zone to comply with time windows)
• Ensuring performance requirements: <1.5 seconds per text, up to 200 simultaneous dialogues​
• Participation in the design of the voice module (ASR/TTS + SIP, latency ≤2.5 seconds) 
• Development of quality metrics: LSS ≥ 99.5%, the proportion of escalations to the operator < 30%, uplift recovery rate ≥ 8%

What we offer (Social package, benefits)

• Format: project employment, remotely
• Duration: Stage 1 — 2026 (pilot for 10,000 debtors), further possible continuation
• Infrastructure: on-premise GPU cluster on customer's equipment (Dell PowerEdge), access to a dataset of 200,000+ records and 50,000+ dialogues

Basic requirements (education, skills)

What you have to do

What we offer (Social package, benefits)

We have our own IT development, which has been operating for almost a year within the same company.

There are no analogues of the development in Kazakhstan yet.

There are plans to enter the Kazakhstani customer market, but the development will need to be finalized.

First of all, senior students in the areas of "Programming" and "Web development" are considered.

Middle, senior, and team-lead specialists)) no need.

It will be necessary to refine the functionality, put it in a web shell / mobile application.

Why are we considering only students? This is a chance to prove yourself and your knowledge not just on paper theory, but in a real business case.

The reward will be provided as the project is completed.

Send your RESUME, portfolio, github, pet projects by email. expertplusdeltam@gmail.com marked "Our project".

Who we are:

Biometric Company.Vision, the Computer Vision team. We make models that “see” the world:

detection/segmentation/classification, data quality is critical.

What needs to be done (the main thing is markup):

● Mark up images for computer vision tasks:

bbox (rectangles)

o Classification / segmentation,

o Key points

o Classes/attributes

Sometimes it's a simple check/correction of already marked up data.

● Work according to clear quality guides and checklists.

● Be careful: the same logic of markup, attention to detail, minimization of errors.

● Keep a small report: progress, questions on controversial cases, suggestions for improving instructions.

In parallel (as development, not instead of markup):

● Dive into the basics of ML/CV: why does the model need such markup, how data affects quality.

● Sometimes participate in mini-reviews of model/dataset errors with the team.

Who we are looking for:

● A diligent and attentive person who likes to do monotonous but important work.

● Willingness to learn and follow markup rules.

● Basic PC confidence, accuracy, responsibility.

● Plus (optional): interest in ML/CV, experience with any markup tools (CVAT, Label Studio, Supervised, etc.).

We offer:

● Full-time internship / part-time (format can be discussed).

● Mentor, clear instructions, quick feedback on quality.

● Real projects and datasets (not “educational markup on the table”).

● Growth: from markup → quality control → more complex markup types and understanding of the ML pipeline.

Biometric.Vision is a leader in biometric identification technologies. We create products that are used by millions of people to access financial and government services. Our success is based on a balance between impeccable security and high conversion rates. We are looking for a Lead of Fraud & Conversion Lab to head our "quality and Safety management center." This is a role for a researcher and strategist who understands that behind every percentage of conversion there is real business money, and behind every case of fraud there are reputational risks.

Your key tasks are:

- Threat Research (R&D): In-depth study of modern methods of bypassing biometric systems (Deepfakes, 3D masks, Injection attacks). You will have to act as the "devil's advocate", finding vulnerabilities in the product before intruders.

- Optimization of the funnel (Conversion): Detailed analysis of the user's path during the verification process. You will need to find the stages at which users "fall off" and offer solutions to improve the UX without losing the quality of security.

- Algorithm setup: Calibrate system response thresholds (FAR/FRR) together with the team, finding the perfect balance for different business segments.

- Analytical reporting: The formation of regular reports for the CEO and key stakeholders on the security status of the system and the effectiveness of customer inspections.

What we expect from you:

- At least 3 years of experience in the fields of Anti-fraud, Risk Management or Product Analytics in large fintech projects or banks;

- High proficiency in SQL and Python (for data analysis and reporting automation);

- Understanding the principles of neural networks and computer vision systems (Computer Vision);

- The ability to think critically and quickly adapt to new types of cyber threats.

What will be the advantage:

- Experience in regulated industries (KYC/AML requirements);

- Mathematical background or specialized education in the field of information security.

Conditions:

- Probation period (3 months);

- Work in the epicenter of AI technologies;

- Competitive salary and the possibility of obtaining options;

- Modern office and a team of professionals.

We are looking for a business/systems analyst who will work on the company's own product in the field of KYC (Know Your Customer), digital identification and compliance solutions for banks, fintech companies and government organizations.

The role involves deep work with requirements, integrations and architecture, close interaction with the product team and development, as well as influence on product development and its value to customers.

Tasks:

- Collection, analysis and formalization of business and system requirements from customers and internal stakeholders;

- Conversion of requirements into User Stories, Use Cases, and Functional Requirements;

- Modeling of business processes and system interactions (BPMN, UML, ERD);

- Development of KYC service integrations: REST API, data formats (JSON/XML), microservices interaction;

- Decomposition of requirements into tasks for backend and frontend developers;

- Checking the completeness and feasibility of requirements together with the development team;

- Participation in grooming sessions, planning, demo and acceptance of functionality;

- Preparation of design, technical and user documentation (including according to GOST);

- Advising team members at all stages of development.

We expect you to:

- At least 3 years of experience as a business or system analyst (IT/ SaaS/ FinTech);

- Confident command of BPMN, UML (sequence, class, component), ERD - required;

- Practical experience in describing requirements: User Story, Use Case, Functional Requirements;

- Knowledge of SQL at the level of writing complex queries for data analysis;

- Understanding the principles of web and mobile application development;

- Experience working with REST API, basic understanding of microservice architecture;

- Experience working with Atlassian stack (Jira, Confluence), Figma, draw.io , miro;

- Understanding and practical experience in Agile/ Scrum/Kanban, knowledge of Waterfall;

- Business communication skills, the ability to negotiate and argue decisions;

- The ability to prioritize tasks and manage time effectively.

It will be a plus:

- Experience working with KYC/AML / Compliance / Identity products;

- Experience interacting with banks, fintech companies, or government systems;

- Experience of integration via Smart Bridge;

- Working with product metrics and user data;

- Google Analytics, Amplitude, Tableau, Power BI;

- Experience in A/B testing of functionality;

- Specialized certifications (Agile, BA, SA);

- English language proficiency.

Conditions:

- Full employment;

- Work schedule: 5/2;

- Work format: office (there is no remote format);

- Work on your own product, not outsourcing;

- Strong product and engineering team;

- The ability to influence architecture, processes, and product development.

We have been developing software since 2015. We develop projects in the B2B segment on various platforms for clients, and develop more than 12 of our own products.

At the moment, we are looking for a product owner with experience in product development, business thinking and understanding of unit economics. Background as a business analyst will be a plus.

The main focus is prioritizing tasks based on value for business and customers. Exactly:

• Decomposition of large tasks.
• Constant analysis of customer feedback and prompt changes.
• Creating clear requirements for features (User Stories, Acceptance Criteria).
• Working with the team: participating in daily stand-ups, sync sessions, incident analysis.
• Prompt response to problematic requests (bug identification, increased convenience).
• Interact with support and the sales department to prioritize user-critical tasks.
• Creation of self-service materials (FAQ, knowledge base) to reduce the burden on support.
• Analyzing customer requests to identify product bottlenecks.
• Testing new hypotheses based on user requests.
• Implementing feedback through built-in tools (surveys, NPS, metrics).
• Collecting information about features on the market and adapting the best solutions.
• Working with developers, UX/UI, QA, marketing, and support.
• Ensuring transparency of the roadmap and decision-making processes.

What we offer:

• A job where every day is a new challenge and new knowledge.;
• An office in which every corner is saturated with comfort: with a stunning view of Bayterek, unlimited coffee and tea, fresh fruits and even its own gym;
• The corporate library;
• 5 bright and unforgettable corporate events per year;
• Improve your language skills through English courses.

• Agile Project Management
• Business analysis
• Unit-economics
• Managing stakeholders
• Roadmap
• Change Management
• Customer Journey Mapping

Our eKYC platform is a high—tech constructor that integrates into the ecosystems of the largest companies. We are looking for a Senior Product Manager who will take responsibility for the evolution of the platform. This is a job for a person who knows how to combine a deep technical understanding of a product with a business vision. You will determine what the biometrics market will be like tomorrow.

Your key tasks are:

- Strategy & Vision: Developing a long-term Product Vision and forming a Roadmap based on market analysis and customer needs.

- Product Delivery: Full lifecycle management of features: from collecting requirements and writing detailed technical specifications to coordinating development (Backend, ML, Mobile SDK) and release.

- Stakeholder Management: Synchronization of expectations between the sales department, marketing, support and the founders of the company.

- Data Driven Approach: Building a system of product metrics, analyzing the use of functions and making decisions based on numbers.

What we expect from you:

- At least 4 years of experience as a Product Manager in product IT (mainly in B2B SaaS, Fintech or CyberSecurity);

- Deep understanding of technical aspects: Rest API, SDK, client-server interaction;

- Experience in managing cross-functional teams in Agile/Scrum methodologies.

What will be the advantage:

- Experience working with products in the field of digital identity or cryptography;

- The ability to conduct high-quality CustDev interviews and identify the true pains of corporate clients.

Conditions:

- Work in the epicenter of AI technologies;

- Competitive salary and the possibility of obtaining options;

- Modern office and a team of professionals.

About the position: The main objective of this role is to minimize the loss of internal expertise and create a single verified source of knowledge. You will become the "keeper of the meanings" of the company, responsible for collecting, structuring and keeping up-to-date documentation.

An important part of the role (about 10% of the time) is assisting the CEO, which will provide you with direct access to decision-making processes at the highest level.

Key responsibilities:

- Knowledge Architecture: Creation and maintenance of a knowledge base structure (Confluence/Notion), providing convenient navigation and search.

- Documentation: Writing regulations (SOP), user instructions and internal manuals "from scratch", communicating with engineers and products.

- Onboarding: Creating training tracks for new employees so that their immersion in the project is as fast as possible.

- Business assistance (10%): Administrative support to the CEO (coordination of meetings, assistance in operational tasks related to business development).

Requirements for the candidate:

- Higher education (management, IT or linguistics).

- English (B2 and above) for working with technical documentation.

- Systems thinking: The ability to transform chaos from the thoughts of experts into a clear and understandable structure.

- High adaptability: Willingness to quickly switch between writing a manual and helping the CEO in an urgent matter.

- Owning Notion/Confluence will be a huge advantage.

Advantages of the position:

- Direct contact with top management: You will be aware of all strategic decisions.

- A deep dive into the AI and Fintech industry at the start of your career.

- International scale: The possibility of business trips around the world (including the countries of Southeast Asia and Latin America).

Conditions:

- Work in the epicenter of AI technologies;

- Competitive salary and the possibility of obtaining options in the future;

- Modern office and a team of professionals.

To respond, send a cover letter telling us about your skills that will help us in our work.

Biometric.Vision is an international company in the field of biometrics and digital identification. We are looking for an intern/intern who will help maintain order in Bitrix24 CRM and conduct research in the following areas: cryptocurrencies, iGaming, banks, telecom.

Suitable specialties:

- Marketing

- Management / Project Management

- Information Systems / Digital Marketing

- Applied Computer Science

- Economics / Finance

- The internship is suitable for students who want to develop in CRM, analytics, digital and research.

What to do:

1. Working with CRM (Bitrix24):

- update transactions, statuses, contacts

- maintain order in the funnels

- check and clean the data

- make simple reports, uploads, tables

2. Research and analytics:

- search for companies and contacts

- explore markets: crypto, iGaming, banks, telecom

- collect data and consider sources

- make mini-analytical notes and compilations

Who we are looking for:

- a 2-4-year student or a young specialist

- attentive, diligent, accurate

- able to work with Google Sheets / Excel

- able to quickly search and structure information

- ready to learn Bitrix24 and analytical tools

- experience is not required - we will teach

It will be an advantage (optional, but it will be a plus):

- Notion / Confluence / Trello

- Excel at the function level (VLOOKUP/XLOOKUP, filters, pivots)

- Google Sheets (formulas, filters, sorting)

- understanding the logic of CRM systems

Interest in:

- cryptocurrencies

- fintech

- IT and digital products

- Biometrics and security

We offer:

- internship for 1-3 months

- flexible schedule (can be combined with studies)

- mentors and training materials

- working experience with CRM and analytics

- the opportunity to receive an offer based on the results