Fractional CTO service and R&D team ramp-up for a pharma startup

How we helped AITHENA go from idea to scalable AI platform with an offshore team.

Industry

Pharmaceuticals

Partnership

2024 - ongoing

Team size

5

Customer’s feedback

Collaborating with OTS has been a seamless experience. Their team delivered a robust MVP, demonstrating genuine adaptability to our feedback. Their professionalism and technical acumen were key to accelerating our early-stage development and client engagement.

Hank Du, Director & Co-Founder

Company overview

AITHENA is a UK-based early-stage startup developing an AI platform to automate the MLR process, helping pharma companies stay compliant with UK and Irish rules while reducing manual work.

Project objectives

The MLR review process is known in pharma for being time-consuming, manual, and expensive. Materials often pass through multiple stages and stakeholders, leading to delays that can span months. AITHENA’s founders came to On The Spot to change that.

Our task

  • Quickly build a secure, scalable MVP to automate parts of the MLR process with AI
  • Evaluate, plan and implement a tech stack ready for fast iterations, regulatory compliance, and future scalability
  • Support the founding team with technical leadership, team hiring, and product strategy

Technology stack

  • Frontend: React.js
  • Backend: Python, PostgreSQL, Redis, Celery
  • AI infrastructure: OpenAI, Azure AI Foundry, LlamaCloud, LlamaIndex
  • Cloud: Kubernetes, Microsoft Azure, Google Cloud

Our solution

What began as a fractional CTO engagement quickly evolved into a close technical partnership. Here’s how our collaboration has been progressing step by step:

  • Technical consulting

    Our fractional CTO has taken ownership of the technical direction, including technology stack selection, system architecture design, and overall technical strategy. We continue to advise on product development, process optimization, and team structure.
  • PoC and MVP development

    In just three months, we built a PoC that proved AI can reliably automate legal reviews. Building on this, we delivered a production-ready MVP — a stable, structured product for early users that we continue to refine and expand.
  • Development team recruitment and management

    Our fractional CTO is directly involved in both product delivery and team building. He is not only engaged hands-on in development on a daily basis, but also advises on team structure and hiring needs as the product evolves.

Technical approach

  • Focus on LLMs and RAG architecture

    We implemented a Retrieval-Augmented Generation (RAG) approach, combining advanced LLMs with smart data retrieval to automate the MLR process. Using frameworks like LlamaIndex and LangChain, we expanded the team with specialists to strengthen expertise.
  • API and external service integration

    Instead of building a complex admin panel, we integrated a no-code database via API, enabling the legal team to manage prompts and data in a familiar environment.
  • Automated testing

    From the very beginning, we embedded automated testing, building a suite of 200+ tests to ensure core features and workflows run reliably.

Results

We helped AITHENA bring their AI platform from PoC to a production-ready MVP, enabling early user testing and client engagement. The solution has already been recognized by adopters as more functional and user-friendly than competitors. Today, we continue to scale the product and team to support AITHENA’s growth.

Real-life team collaboration

To strengthen collaboration, AITHENA’s founders visited our Warsaw office for in-person meetings with the development team. It was a great chance to connect offline and sync up on key topics.

They also joined our TechSpot event as speakers in December 2024 to share their experience of building an AI startup as non-technical founders. You can check the recording here.

This is some text inside of a div block.
Customer visiting the Warsaw team (from left: Vlad Paddubny, William McCully, Hank Du, Anton Vasiljev)
Team-building in Kraków
Hank Du and William McCully at AI Focused TechSpot

Ready to build your R&D team in Poland?

contact us
contact us