Projects

Resume

Projects

Resume

Home

HAZOP Software

Soter Safety Consulting

Designing a B2B SaaS software for process manufacturing industries breaking into a $1.3bn market still running on legacy tools built in the early 2000s

A HAZOP is a structured safety review used to catch hazards in complex industrial processes before they become real problems, run across process manufacturing industries. The existing legacy platform forced process safety engineers and its multiple stakeholders through a data structure built for storage, not judgement, burying the terminology, traceability and progress signals a regulated, data-dense, risk-assessment workflow depends on.

I led UI/UX across the end-to-end design of a tiered, multi-user enterprise SaaS product: interaction model, navigation architecture, design system and build, presenting design decisions directly to client stakeholders and contributing to feature scoping and the launch roadmap as part of the broader GTM strategy.

My Role

Product Designer

Team

6 Designers

1 Product Manager

Client stakeholders

Timeline

5 Months

Outcome

What changed?

The build matched how engineers actually reason through a HAZOP study. A progressive disclosure model replaced the rigid, linear flow, surfacing only what was needed, when it was needed, and pairing the worksheet with a live P&ID view so engineers never lost context switching screens. This cut error in a workflow where mistakes carry regulatory weight, and gave Soter a design foundation flexible enough to scale into future Process Risk Study tools.

Improvement in task completion

Reduction in user error

Improvement in platform consistency

My Role

What did I contribute?

I was the UI/UX Designer responsible for end-to-end product design across heuristic evaluation, IA restructuring, usability testing, design system and platform build for a two-tier B2B SaaS product (Facilitator and Enterprise). This was a client-facing engagement where I wore multiple hats.

Research

Voices from research

Two rounds of qualitative research, stakeholder interviews and a 21-person pilot survey, shaped the redesign, and kept shaping it as testing continued post-launch.

Engineers couldn't trust the system

"The user interface is un-engaging with limited functionality to generate dashboards."

Interviews with a facilitator and a senior Risk Assessment consultant surfaced the same core complaint from two different vantage points:

Terminology didn't match their Excel workflows
There was no progress indication, no way to go back
Every step risked losing safety-critical data.

One interface couldn't serve every role

"The desire for an improved user interface and navigation between nodes and keywords... option to mark critical items."

At least 7 distinct roles touch the same underlying dataset across the HAZOP lifecycle, from facilitator through to external stakeholders. Each has a different reason to be in the tool and a different task shape.

The pilot study backed this up at scale: 62% of the 21 respondents already used HAZOP software, yet 85% still wanted a new tool, citing weak navigation and no way to view HAZOPs without horizontal scrolling.

Excel wasn't a competitor, it was a safety net

"Exporting data to Excel, as it's widely used in companies, ensuring a smooth transition."

Both interviewees said exporting to Excel was essential, and 63% of Excel-user survey respondents agreed.

This wasn't about nostalgia for Excel, it was the only reliable way to guarantee adoption across companies running different HAZOP formats.

Insights

What was broken?

A systematic heuristic evaluation against Nielsen's 10 principles across every screen, cross-checked against usability testing findings, surfaced issues across 4 areas:

Projects Page: No way to edit, archive or filter projects. Profile icon merged with software name
HAZOP Preparation: No progress indicator, Terminology mismatch with Excel, No way to undo
Data Entry Page: Prematurely active add buttons. No error messaging. Scenario navigation unclear
Analytic Dashboard: No project name shown. Charts unlabelled. No user control over displayed data

A complete audit and research of the website focused on 4 areas mainly:

Sports Page: Entry point for majority of the interactions
Match Page: Specific to MENA region, High priority for Euro 2024
General Audit: Includes all pages, sports, languages & devices
Design System: Barely functional

(Hover on the numbers for more info)

Projects

HAZOP Preparation

Data Entry

Analytics Dashboard

The industry itself hadn't moved since 2010: paper, then word processors, then Excel and Access-era tools, while software everywhere else went through a fourth transition to intuitive interfaces, cloud storage and analytics. The cost was measurable, not abstract, up to 30% of HAZOP time lost to manual data handling, industry-wide.

These insights were connected to the metrics being affected giving a clear direction to work on.

Terminology Mismatch

Fields didn't match the Excel sheets engineers already worked from, and progress signals were missing throughout

Increasing error rate & lowering user trust in the platform

Rigid, One-Size-Fits-All Screens

Facilitators, Project Leads and Senior Management all needed different views, but every screen treated them the same

Negatively impacting task completion rate

Unclear Task Flow

Prematurely active buttons, no error messaging, and unclear scenario navigation slowed engineers down mid-task

Increasing time-on-task & drop-off during data entry

Key Design Decisions

What I along with others changed?

Building an operator-centric model

PRESSURE

The starting skeleton mirrored Soter's category logic, not how engineers actually work.

CALL

Rebuilt the interaction model around real HAZOP reasoning, keeping the skeleton intact underneath.

RESISTANCE

The client knew the skeleton well; changing it meant justifying every shift in a risk-averse, rigid & highly regulated industry.

RESULT

Task completion up 35%, errors down 28%, on a platform where mistakes carry regulatory risk.

Trading broader scope for a leaner, stable core

PRESSURE

Early pressure to build toward LOPA/PHA fast and lock users into a rigid wizard flow.

CALL

Prioritised the core HAZOP workflow first, using progressive disclosure and role-based views instead, validated by the client's own scope-narrowing to their MVP.

RESISTANCE

Stakeholders wanted visible progress on the full PSM vision sooner, not a narrower first release.

RESULT

Fewer steps without losing structure, setting up a 3-year roadmap built on a proven core.

Aligning domains through testing and stakeholder sign-off

PRESSURE

Oil & Gas and Pharma stakeholders assumed their requirements needed separate configurations.

CALL

Ran cross-functional workshops and guerrilla usability testing, then presented evidence-backed rationale to leadership.

RESISTANCE

Winning consensus meant leadership had to trust the testing evidence, not just the designer's word.

RESULT

One interaction model for both domains, lifting platform consistency by 40%.

Beyond the brief

Alongside the platform build, the team also shaped Soter's go-to-market path:

Go-To-Market(GTM) strategy and competitive benchmarking

5-year roadmap sequencing HAZOP as the stable core, expanding into LOPA, PHA and other process safety tools
A clear path to 5% of the UK HAZOP market within 5 years

Other Projects

beIN Sports

End-to-end redesign for the MENA region's leading live, high traffic multilingual sports platform - web & mobile. Resulting in +518% engagement & 105% site retention.

TATA Play - Hybrid IPTV

A Product as a Service(PaaS) IPTV framework unifying 27+ OTT platforms, Live TV and VOD for 17+ Indian language audience

Mind Way

Designing an ecosystem for the neurodivergent where inclusion is given, and not an afterthought

Back

deepikaraju09@gmail.com