Optimization and Machine Learning for SMEs

Goal

Modeling, Analysis, Optimization — All from One Source

Atlas Optimi­zation GmbH was foun­ded in 2021 with the aim of making the cur­rent achie­ve­ments in mathe­ma­ti­cal optimi­zation and machi­ne lear­ning acces­si­ble to small busi­nesses. Lar­ge cor­po­ra­ti­ons main­tain their own depart­ments with spe­cia­lists who mathe­ma­ti­cal­ly for­mu­la­te and sol­ve real-world pro­blems, thus con­tri­bu­ting to the impro­ve­ment of pro­ducts or the auto­ma­ti­on of ope­ra­tio­nal workflows.

The meta-stu­dy focu­sed on logi­stics appli­ca­ti­ons [1] esti­ma­tes that such opti­miza­ti­ons have led to effi­ci­en­cy increa­ses of on avera­ge 20 — 30% or have intro­du­ced enti­re­ly new approa­ches to solu­ti­ons. The­se advan­ta­ges are usual­ly not available to small and medi­um-sized enter­pri­ses. We want to chan­ge that.

Competence

Optimal Decisions under Uncertainty

For­mu­la­ting and sol­ving real-world pro­blems mathe­ma­ti­cal­ly is noto­rious­ly dif­fi­cult, but as for­mer rese­ar­chers at ETH Zurich, we are experts in this area. Whe­ther it invol­ves the opti­mal esti­ma­ti­on of unknown quan­ti­ties, the sel­ec­tion of sys­tem and con­trol para­me­ters, or the repli­ca­ti­on of com­plex input-out­put rela­ti­onships — deri­ving opti­mal decis­i­ons under uncer­tain­ty is our spe­cial­ty. In this pro­cess, we uti­li­ze sta­te-of-the-art methods of appli­ed mathe­ma­tics to for­ma­li­ze your issue as an optimi­zation pro­blem in a line­ar, high-dimen­sio­nal space. Typi­cal­ly, the solu­ti­on pro­cess its­elf is not a pro­blem. Even with free­ly available open-source soft­ware, optimi­zation pro­blems with hundreds of thou­sands of varia­bles can be relia­bly sol­ved. A con­stant­ly updated list of exem­pla­ry appli­ca­ti­ons can be found here.

Offerings

Have you encoun­te­red a pro­blem in your work that is likely mathe­ma­ti­cal in natu­re? A pro­blem whe­re decis­i­ons need to be made despi­te not having access to all the infor­ma­ti­on? Just cont­act us. You can expect straight­for­ward inter­ac­tions and a serious inte­rest in your case from us. We are hap­py to advi­se you on the mathe­ma­ti­cal mode­ling of your pro­blem, pro­vi­de a solu­ti­on, train your employees, and offer tuto­ri­als or even initia­te a rese­arch pro­ject with you.

I’m loo­king for­ward to it!

Dr. Sc. ETH Jemil Avers Butt, CEO

Sources

[1] C. For­za & F. Di Nuz­zo (1998) Meta-ana­ly­sis appli­ed to ope­ra­ti­ons manage­ment: Sum­ma­ri­zing the results of empi­ri­cal rese­arch, Inter­na­tio­nal Jour­nal of Pro­duc­tion Rese­arch, 36:3, 837–861

Vision

We make your life easier. Sta­te-of-the-art optimi­zation algo­rith­ms gua­ran­tee you the best pos­si­ble results for a wide ran­ge of decis­i­on-making problems.

We solve

Values