See notes from Ivorra talk Oberwolfach 2009
http://ncatlab.org/nlab/show/A-infinity-category
http://mathoverflow.net/questions/815/triangulated-vs-dg-a-infinity
The following paper by Roitzheim and Whitehouse was withdrawn due to some errors: Abstract: This paper investigates if a differential graded algebra can have more than one -structure extending the given differential graded algebra structure. We give a sufficient condition for uniqueness of such an -structure up to quasi-isomorphism using Hochschild cohomology. We then extend this condition to Sagave’s notion of derived -algebras after introducing a notion of Hochschild cohomology that applies to this.
Batanin-Cisinski-Weber on some globular operads stuff that in particular gives a tensor product for A-infty algebras
arXiv:1102.1311 Interchanging A_\infty and E_n structures from arXiv Front: math.AT by Zbigniew Fiedorowicz, Rainer M. Vogt The notion of interchange of two multiplicative structures on a topological space is encoded by the tensor product of the two operads parametrizing these structures. Intuitively one might thus expect that the tensor product of an E_m and an E_n operad (which encode the muliplicative structures of m-fold, respectively n-fold loop spaces) ought to be an E_{m+n} operad. However there are easy counterexamples to this naive conjecture. In this paper we show that the tensor product of a cofibrant E_m operad and a cofibrant E_n operad is an E_{m+n} operad. It follows that if A_i are E_{m_i} operads for i=1,2,…,k, then there is an E_{m_1+m_2+…+m_k} operad which maps into their tensor product.
nLab page on A-infinity stuff [private]