بهین بایگان داده هونام | Behin Baygan Dadeh Hoonam

مخابرات و ارتباطات (تلکام) ۱۱ تا ۵۰ نفر siz-tel.ir/

فرصت‌های شغلی ۳

استخدام Ai engineer

این آگهی منقضی شده است

دسته‌بندی شغلی

وب،‌ برنامه‌نویسی و نرم‌افزار
موقعیت مکانی

تهران ، تهران
نوع همکاری

تمام وقت
حداقل سابقه کار

کمتر از سه سال
حقوق

توافقی

شرح موقعیت شغلی

bout the Role

We are looking for a highly skilled and forward-thinking AI Engineer to join our growing team. In this role, you will work at the intersection of language models, retrieval systems, OCR, and infrastructure — helping us build and deploy open-source, self-hosted AI systems that operate at scale and in secure environments.

This is not a research-only role — it’s hands-on, applied, and impactful. You'll design intelligent components, work with air-gapped GPU clusters, collaborate with MLOps to operationalize models, and solve real business problems such as Persian handwriting recognition, multilingual search, and domain-specific retrieval. If you're passionate about cutting-edge open-source AI, multilingual NLP, and real-world deployment, we want to hear from you.

What You’ll Be Doing

As an AI Engineer, you will:

- Design and develop Retrieval-Augmented Generation (RAG) pipelines using open-source tools such as LangChain, LlamaIndex, and vector search technologies like FAISS, Qdrant, or Weaviate to power intelligent assistants and search systems.

- Improve and fine-tune open-source embedding models for semantic retrieval — including adaptations for Persian-language and multilingual content.

- Customize OCR models to enhance their ability to accurately recognize Persian handwriting, particularly for use in scanned documents or handwritten forms.

- Adapt and fine-tune LLMs (e.g., LLaMA, Gemma, DeepSeek) to internal datasets or industry-specific tasks using methods like LoRA, QLoRa, SFT, and instruction tuning.

- Deploy LLMs in secure, air-gapped GPU clusters using tools such as Ollama, vLLM, or SGLang, ensuring low-latency and high throughput private inference in isolated environments.

- Integrate models with real-time user interfaces such as Open WebUI or custom-built frontends, making LLM capabilities accessible to users and teams.

- Develop intelligent agents capable of memory tracking, step-by-step reasoning, task delegation, and contextual decision-making in long or complex interactions.

- Collaborate closely with our MLOps team to ensure models are containerized, monitored, and deployed consistently — with attention to reproducibility, versioning, and performance.

- Continuously explore, evaluate, and integrate emerging open-source tools, benchmarks, and model architectures to ensure our stack stays state-of-the-art.

What You Bring

To succeed in this role, you should have:

- At least Bachelor’s degree in Computer Science or Computer Engineering.

- At least 2 years of hands-on experience in Machine Learning, Natural Language Processing, or building and deploying LLM-based systems.

- Strong skills in Python and practical experience with Hugging Face Transformers, LangChain, or LlamaIndex.

- Experience deploying models using tools like Ollama, vLLM, or SGLang in On-premises GPU environmen

- Familiarity with vector-based retrieval systems and the ability to improve embedding models for multilingual domains.

- Working knowledge of OCR systems, especially related to handwriting or non-Latin scripts like Persian.

- Ability to integrate models into frontend systems, APIs, or internal applications.

- A proactive mindset when it comes to learning and applying the latest advancements in open-source AI.

Bonus Qualifications (Nice to Have)

- Knowledge of quantization, model optimization, or ONNX conversion for fast inference.

- Contributions to open-source projects in the LLM, OCR, RAG, or NLP ecosystem.

- Experience designing or working with multi-agent systems, session-based memory, or planning modules.

- Hands-on involvement in Persian-language AI projects or domain-specific document processing.

Why You’ll Love Working With Us

- Real infrastructure: You’ll work on AI systems deployed on secure, large-scale High resource GPU clusters — not just demos projects.

- Open-source-first: We build with and contribute to open-source technologies. You’ll never be stuck behind a vendor lock-in.

- Cross-functional collaboration: Work closely with MLOps, product, and backend teams to build robust,

maintainable systems.

- Meaningful work: From multilingual retrieval to local handwriting OCR, your work will have practical and high-impact use cases.

- R&D freedom: You’ll have the autonomy to explore and adopt the best tools, models, and practices from the ever-evolving AI landscape.

ثبت آگهی استخدام در جابینجا

معرفی شرکت

شرکت دانش بنیان بهین بایگان داده هونام در سال ۱۳۹۶ با هدف طراحی؛ توسعه و ساخت تجهیزات مخابرات سلولی نسلهای مختلف تشکیل گردید.
این شرکت موفق شده است در طول این مدت با جذب و همکاری متخصصین خبره حوزه سلولار به فناوری های نوینی همچون طراحی و ساخت سامانه LTE تاکتیکی؛ سامانه BBU بومی؛ سامانه هسته بومی مخابرات سلولی نسل 4 و 5 مبتنی بر NFV با پشتیبانی از یک میلیون کاربر و نیز سامانه IMS برای سرویس های VxLTE/VxNR دست یابد.
همچنین یکی از نقاط قوت این شرکت پیاده سازی و توسعه سرویس های مانیتورینگ مختلف و جامع به منظور نظارت بلادرنگ بر سامانه ها و شبکه های مختلف توسعه داده شده در سطح ملی می باشد.
یکی از اهداف اصلی این شرکت پیشروی و پیشبری کشور در حوزه بومی سازی تجهیزات مخابرات سلولار با تکنولوژی سطح بالا و همچنین رقابت پذیر با محصولات عرضه شده توسط شرکتهای مطرح دنیا است.