In this work, the discoloration of azo pigment Eriochrome Black T (EBT) was analysed using a complete methodology consisting of: i) a hydrodynamic study of the photochemical reactor employed for discoloration; ii) experimental analysis of the efficiency based on Design of Experiments (DOE) approach; iii) efficiency modelling using Response Surface Method (RSM); iv) process optimization based on RSM and Harris Hawk Optimization (HHO) algorithm. The reactor has a UV-C lamp protected by a quartz tube and was analysed using the tracer method. The Residence Time Distribution (RTD) technique employed used as tracer a sodium chloride solution of 10% mass concentration. The response was registered with a WTW conductivity meter Cond 315i. The classic RTD functions, namely: residence time distribution I, cumulative residence time distribution (F), internal age distribution (I) and intensity function (λ) were plotted as function of dimensionless time (θ), in normalized form. Also, the first, the second, the third and the fourth moments of residence time distribution, usually known as the mean residence time (η = 14.15), the variance of distributions (ζ = 5.81), the skewness (s = 6.53) respectively kurtosis (k = 4.08) were determined for debit of 28.99 mL/s. The obtained results indicated the presence of flow short-circuits and dead-zones. The next step of the applied methodology considered in an experimental study regarding the EBT discoloration based on hν TiO2. To this mean, the DOE strategy was applied taking into account the following parameters: the amount of photocatalyst (Z1, g/mL), the contact time (Z2, min) and the initial concentration of the dye (Z3, mg/L). Based on the experimental data previously obtained and the RSM based approach, the regression equation of the process was determined in Minitab software (Eq. 1). The variance of the model was 97.76%, indicating a good agreement between experimental and predicted. Y = 6.41 + 17.92*Z1 + 0.0314*Z2 – 0.763*Z3 – 7.41*Z1*Z1 + 0.000003*Z2*Z2 + 0.00661*Z3*Z3 + 0.00325*Z1*Z2 + 0.0029*Z1*Z3 – 0.000338*Z2*Z3 (Eq.1) The last step consisted in the application of a process optimization procedure. In this case, two distinct strategies were applied: i) the RSM approach from the Minitab software and ii) the HHO algorithm (a bio inspired metaheuristic that simulates the cooperative behaviour and chasing style of Harris‘ hawks). The results showed that through optimization, the process efficiency was raised from 75.35% to 98.37%, indicating the capacity of the applied methodology to improve the process without the need for additional resources. Acknowledgements: This work was supported by ―Program 4. Fundamental and frontier research. Exploratory research projects‖ financed by UEFISCDI, project no. PCE 58 2021.