Can we retrieve balances?
A returned balance verifies read access.
From fragmented trading tools to an institutional client experience
I connected the transparency positioning to a client-facing monitoring experience, alongside unified operational tools and new service journeys as CLS evolved toward Motion Trade.
THE CHANGE
Before. Scattered operational tools and engineer-facing dashboards.
Why change. Institutional clients needed the transparency promise demonstrated in the product.
My contribution. Led operations design, initiated client monitoring and reshaped service journeys.
Outcome. Direct client visibility, greater account-manager capacity and a new service architecture.
01 / OPERATIONS
CLS provides market-making and liquidity services. Client records, bot states and balances were scattered across tools; clients relied on account managers to interpret technical dashboards.
I joined as the sole designer and led the admin-platform architecture with engineering and operations. Later, I initiated client monitoring alongside the institutional repositioning.
DESIGN DECISION / ACCESS
A client record does not prove that an API key can read balances or place trades. I separated setup, read-access validation and trade-access validation, with feedback for each check. A test order and cancellation confirm trading access before bot configuration; missing CRM records have a fallback path.
CLS · Exchange permissions
The operator verifies two distinct capabilities before configuring a trading bot.
A returned balance verifies read access.
The check reports the test order and its cancellation.
The operator has explicit feedback about read and trade access.
These are separate checks: a successful connection or balance read does not imply trading permission.
DESIGN DECISION / RECOVERY
I kept error explanations, restart actions, logs and support guidance beside the client, account and balances. Operators could investigate and recover a bot without losing its financial context.
CLS · Order entry
I separated single-order entry from bulk setup so each form shows only the controls needed.
Set up one buy or sell order.
Set price ranges and order counts for buys and sells.
After submission, a confirmation links to the order book to review the result.
02 / CLIENT MONITORING
Institutional positioning promised transparency. I made it tangible through client-facing monitoring, replacing engineer-oriented Grafana views with a dashboard clients could interpret themselves.
Client questions and sales needs shaped the hierarchy: performance, liquidity and system health. I distinguished market movements from committed funds and made exchange, pair and time scope explicit. The web/mobile experience connected with Telegram; mobile filters retained context across views.
CLS · Client monitoring
Three complementary mobile views keep exchange, pair and account context visible.
Swipe to explore the full diagram
Review price, volume and the selected service.
Change exchange and pair in a dedicated view.
Inspect balances, capital in use and liquidity.
Each view answers a different financial question within the selected market context.
OPERATIONAL OUTCOMES
Capacity and inquiries: roughly one month after monitoring launch, including training. Bot recovery is a separate measure. Approximate recalled outcomes of the wider team’s work.
03 / INSTITUTIONAL SHIFT
Alongside monitoring, I led the website and visual direction as CLS evolved toward Motion Trade. This was a repositioning: the core market-making business remained unchanged.
Five client interviews, competitor analysis and Business Development input shaped three situations: launch preparation, liquidity growth and exchange pressure. I restructured 37+ pages, dashboards, cases and infographics around client concerns, evidence and a next step.
Marketing, Business Development and Product assessed clarity and usefulness in sales conversations. Test demos received positive feedback; the earlier conversion figures below do not measure this repositioning.
TEAM LEADERSHIP
I joined alone, hired four designers and led the five-person team while staying hands-on. With HR, I created development plans for every designer and held regular one-to-ones.
I aligned sprints with Product and Engineering, reserved 20% of capacity for unplanned work and built shared design patterns.