Using top-electrodes, we demonstrate the soliton-based miniaturized integration of electro-optic devices in a photorefractive paraelectric bulk crystal. Self-trapping and beam manipulation though soliton electro-activation is achieved at quasi-digital voltages. Photorefractive spatial solitons [1] can be implemented to steer beams [2] and to achieve electro-optically activated beam manipulation functions, such as intensity modulation and two-channel routing, the devices themselves self-integrated in the bulk [3,4]. The optical circuitry is at once integrated in the bulk crystal volume, transparent at infrared wavelengths, and compatible with a fast response.To date, no physical mechanism has been devised to allow for the possibility of imprinting a given steering pattern in the volume, formed through different localized soliton processes, and to subsequently address and electro-optically activate them through appropriate control signals, involving accessibly low switching voltages. We here demonstrate a viable avenue based on delivering the bias voltage through a single facet geometry, with all the electrodes deposited side-by-side, during the nonlinear propagation writing stage and in the subsequent routing stage [5]. Electro-statically, in this configuration, the values of external bias required to achieved kV/cm fields can be considerably reduced by microscopically reducing the electrode distance. Furthermore, by tailoring the electrode geometry, we can achieve a quasi-arbitrary spatial pattern of the bias electric field, up to several hundred microns inside the sample.We implement the geometry illustrated in Fig.(1a), which constitutes, in our scheme, the basic building block to the more complex and elaborate addressable photonic array [6,7,8]. We carry out experiments, in a 3 (x) x2.4 (y) x1 (z) mm sized sample of Cu and V co-doped potassium-lithium-niobate-tantalate (KLTN) crystal, at a temperature T =16 • C, 4 degrees above the ferro-paraelectric transition, where the index of refraction is n 2.4, the quadratic electro-optic coefficient g 0.12m 4 C −2 , and the relative dielectric constant r 1.5 · 10 4 . The top-sided electrodes were tailored on the 3 (x) x1 (z) mm facet, with an intra-electrode 180 µm gap (see Fig.(1b)). The beam was a continuous-wave Argon ion 50 mW x-polarized laser, operating at λ=514nm whose intensity distribution is approximated by a Gaussian of intensity Full-Width-at-HalfMaximum (FWHM) ∆x = ∆y = 7µm and intensity ratio between the maximum beam intensity and the homogeneous background illumination I max /I b = 5.Experimental results for a launch approximately 50 µm from the top edge, are reported in Fig.(2). For a zero applied voltage, the beam diffracts to 15 µm after the 1 mm propagation along the z axis, corresponding to approximately 2 diffraction lengths (see Fig.(2b)). As predicted through numerical simulation, two-dimensional self-trapping is achieved for a V sol 40-50 V, after a build-up time of approximately τ s 120s, for a 1 µW launch.Similar self-trapping pheno...