In machining processes, cutting forces measurement is essential to allow cutting process and tool conditions monitoring. Moreover, in order to have information about the quality of the milled part, the amplitude of the tool tip vibration would be very useful. Since both the measurements are extremely complicated especially in an industrial scenario, in this study, an in-process model based estimator of cutting forces and tool tip vibration was designed and properly tested. The developed estimator relies on both a machine dynamic model and on indirect measurements coming from multiple sensors placed in the machine. The machine dynamic model was obtained through an experimental modal analysis session. The estimator was developed according to the Kalman Filter approach. The fusion of multiple sensors data allowed the compensation of machine tool dynamics over an extended frequency range. The accuracy of the observer estimations was checked performing two different experimental sessions in which both the force applied to tool and the tool tip vibration amplitude were measured. In the first session, the tool was excited with different sensorized hammers in order to appreciate the broad-bandwidth of the performed estimations. In the second one, real cutting tests (steel milling) were done and the cutting forces were measured through a dynamometer, tool tip vibrations were measured as well. The experimental results showed that the indirect estimation of cutting forces and tool tip vibrations exhibit a good agreement with respect to the corresponding measured quantities in low and high frequency ranges. The contribution of this research is twofold. Firstly, the conceived observer allows estimating the tool tip vibrations that is a useful information strictly connected to the surfaces quality of the processed workpiece. Secondly, thanks to a multi-sensors approach, the frequency bandwidth is extended especially in the low frequency range.