I will fine tune, evaluate or deploy llms for your business


About this gig
Custom LLM & NLP Solutions Fine-tuning, Evaluation, Deployment
Are you looking to leverage Large Language Models for your business but not sure where to start? Or do you have an existing LLM application that needs evaluation and improvement?
I specialize in practical, productionready NLP solutions for European businesses. I'll work with you to understand your use case and deliver a solution that actually works.
What I can help you with:
LLM Finetuning Adapt opensource models (Llama, Mistral, Gemma) to your specific data and domain.
LLM Evaluation Measure accuracy, hallucination, latency, and cost; compare multiple models
Document Processing Extract, summarize, or classify information from PDFs, emails, or websites.
Custom Chatbots Build a chatbot for your internal knowledge base or customer support
Text Classification & NER Categorize support tickets, extract entities from legal/medical texts, etc.
Sentiment Analysis Understand customer feedback at scale.
My toolkit: Python, PyTorch, Hugging Face, Transformers, spaCy, NLTK, FastAPI, LangChain (if needed)
I am available for European business hours. Click the contact button and tell me about your project!
Get to know Tahmina A
AI and Machine Learning Engineer
- FromDenmark
- Member sinceMar 2026
- Avg. response time2 hours
- Last delivery1 day
Languages
English, Bengali, Danish
My Portfolio
FAQ
What exactly do I get with each plan?
Please see the pricing plans and it's description.
Do you work with my own data?
Absolutely. You provide your dataset (CSV, JSON, Excel, text files, or PDFs), and I'll work with it. For sensitive data, we can sign an NDA if needed. All data is handled confidentially and never shared.
What about GDPR? I am based in Europe.
Great question. I am based in Denmark and fully respect GDPR requirements. I do not store your data after delivery unless agreed otherwise. No data is sent to third-party APIs without your explicit permission. If you need open-weight models only (no OpenAI/Anthropic APIs), I can accommodate that.
Do you use OpenAI / ChatGPT or open-source models?
I primarily use open-source models (Llama, Mistral, Gemma, Phi) via Hugging Face and PyTorch. This means your data stays private, no API costs, and full ownership of the fine-tuned model. If you prefer OpenAI/Anthropic APIs, I can also work with them — just let me know.
Can you deploy the model for me?
Yes. The Premium plan includes FastAPI deployment ready for your environment (local, cloud, or your own server). I will provide clear instructions and can help with basic setup. For full production deployment on AWS/Azure/GCP, that can be arranged as a separate order.
What languages do you support?
English first. For other European languages (Danish, German, French, Spanish, Italian, Dutch, Swedish, Norwegian), most open-source LLMs handle them well. I can test and advise based on your specific language needs.
How do we get started?
Simply message me before ordering. Tell me a bit about your use case, data format, and expected volume. I'll confirm which plan fits best and answer any questions. Then you place the order and share your data. I deliver within the agreed timeline.
What if I need ongoing support after delivery?
The Premium plan includes one week of post-delivery support. For longer-term maintenance, new features, or additional fine-tuning, I'm happy to discuss a separate ongoing arrangement or a new order.
What if I am not satisfied with the delivery?
Your satisfaction is important to me. Each plan includes a set number of revisions. If you feel the work doesn't meet the agreed scope, I will work with you to make it right before marking the order complete.
Do you offer consulting or advice before I order?
Yes, I offer a free initial consultation (15–20 minutes via Fiverr chat or call) to understand your project and recommend the right approach. No obligation to order afterward.

