top of page
Neutral Hoodie Portrait_edited_edited.jpg

RandyKeeler

Randy Keeler

PROJECT

Atlas

AI Platform

Overview: Develop an AI Platform

Atlas is a proprietary internal AI platform designed to give employees self-service access to organizational knowledge and generative AI — entirely within a secure, company-controlled environment. Purpose-built to handle competition-sensitive data that could never be trusted to commercial tools, the platform combines a graph database, a conversational AI interface, and a suite of custom workflow apps. Randy Keeler led the initial stages of the version 1.0, UI/UX design effort, shaping a cohesive product experience that spans proposal content generation, HR self-service, past performance retrieval, and capability mapping.

CloseUp-05.jpg



Making the Complex Simple

Atlas combines multiple AI components, LLM data models, and workflow tools under one roof — and the central design challenge was making all of that complexity invisible to the end user. If the interface felt technical or required any ramp-up, adoption would stall regardless of how powerful the platform was underneath. Compounding this, many users would spend extended hours on the platform daily, making visual comfort and ergonomics an equally important design constraint alongside usability.

The Challenge

Extracted fromAtlas 1228 - Mgr Added.jpg



Familiar by Design. Powerful by Choice

Randy designed the initial version 1.0 interface around conversational UI conventions familiar from consumer AI products, so users felt oriented from their very first login. Advanced capabilities — model selection, prompt libraries, graph filtering — were surfaced as secondary controls shown only when needed, keeping the experience approachable without limiting depth for power users. A muted, cool-toned color palette was selected to reduce eye fatigue across long sessions, with a calm, consistent visual language applied across every screen of the platform.

The Solution

CloseUp-06-Thumbs.jpg



 

From Days to Hours with AI

The design paid off where it counts: adoption and speed. Atlas launched to the full employee base in 2024 with no formal training required, and employees company-wide reported faster daily workflows, some measuring improvements of up to 47%. Proposal teams felt it most: RFIs and RFPs that previously took days to assemble were going out the door in a few hours.

These results aren't Randy's claims alone. In a company podcast series he created to drive Atlas adoption, users described in their own words how the platform saved them time and contributed to winning client contracts. Designing the product and then producing the channel that proved its impact is the kind of full-loop ownership Randy brings to every project.

The Results

A Glimpse of Outcomes

47%

Faster daily workflows reported after launch

Hours vs days

RFP & RFI Turnaround

1,425+

Employees Onboarded Company-wide
freepik_place-image-into-a-deskto_2810330767.jpg

Conclusion

What previously took days or weeks to produce, can now be completed in hours, without being limited by resource availability.
 

bottom of page