In the field of artificial intelligence systems development, Akash, co-founder and CEO of Bellum.ai, has positioned himself as a relevant figure by addressing the challenges and best practices for implementing this technology in companies. During a recent talk, Akash shared his journey and the lessons learned from working with numerous organizations that have integrated AI into their operations. His purpose is to clarify which strategies are effective, which are not, and where artificial intelligence is heading.
Akash’s relationship with artificial intelligence began approximately four years ago, coinciding with the outbreak of the COVID-19 pandemic, when he began experimenting with the GPT-3 API. Despite the limitations of this technology, which produced some inaccurate or random responses, it also demonstrated a surprising ability to complete phrases coherently. During this period, his work in recruitment software allowed him to apply AI to creating job descriptions and classifying emails, leading to the viralization of his AI-powered job description generator, showcasing the potential for automation through artificial intelligence.
However, Akash emphasized that implementing these models in production environments posed significant challenges. Aspects such as engineering prompts and collaboration in workflows were complex to manage. The situation took a turn with the launch of ChatGPT in November 2022, which propelled AI into the mainstream. The obstacles Akash had encountered in using AI now extended to numerous industries, highlighting an urgent need for effective solutions.
Aware of this new reality, Akash and his co-founders established Bellum.ai, with the mission of empowering companies to maximize the use of advanced language models (LLMs) and develop robust AI systems. Akash’s experience at McKinsey provided valuable insights into AI governance and the evolution of related technologies. The growth of GPT models and their increasing impact across various sectors strengthened his conviction about the need for structured frameworks for the safe and effective deployment of artificial intelligence.
via: MiMub in Spanish
