Network slicing is a key feature of 5G and beyond networks. Intelligent management of slices is important for reaping its highest benefits which needs further exploration. Focusing only on one goal as revenue maximization or cost minimization may not generate the highest profit for infrastructure providers in the long run. In this paper we jointly consider online admission and placement of Radio Access Network (RAN) slices with two objectives-a) maximizing revenue from accepting slices which are more profitable in the long run, and b) minimizing the cost to deploy them in Open RAN (O-RAN) enabled network by placing the slices efficiently. We formulate it as an optimization problem and propose a Deep Reinforcement Learning (DRL) based solution using Proximal Policy optimization (PPO). We compare our model with a state-of-the-art DRL based admission control solution and a greedy heuristic. We show that our proposed solution can efficiently adapt to dynamic load conditions. We also show that the proposed solution results in better performance to maximize the overall profit for infrastructure providers in comparison to the baselines. © 2022 IEEE.