The product has outgrown the first version.
Core workflows are brittle, the data model no longer matches the business, or every release exposes another hidden dependency.
Your hardest software problem. Diagnosed, built, and shipped.
I join your team as a hands-on principal engineering partner. I diagnose the problem, shape the system, ship it into production, and leave your team able to own it.
Product engineering, software and platform engineering, and AI engineering for growing technology companies.
- 15+ years
- Hands-on production engineering
- Shopify, Brex, Motorola
- Career experience inside large production systems
- Direct
- The person shaping the work also ships it
When companies call Modh.
The common thread is consequence. The work crosses product, architecture, and operations, and it needs one person to carry the whole problem.
A consequential build has no clear technical owner.
The company knows the outcome it needs, but product scope, architecture, delivery sequence, and production responsibility are split across people.
The team is shipping, but the system is fighting back.
Migrations stall, incidents repeat, integrations disagree, performance slips, or engineers cannot change one area without breaking another.
AI needs to do real work, not perform a demo.
The model must operate inside permissions, data boundaries, evaluations, human review, failure handling, and a workflow the business can trust.
The software is the proof.
Real products, live operations, and technical decisions that had to hold up after launch.
AI sales performance platform
A revenue operating system built around the real sales workflow
A production platform connecting lead capture, qualification, scheduling, CRM activity, sales outcomes, attribution, billing, and operational reporting.
- The problem
- Sales activity, bookings, CRM outcomes, attribution, and billing had to agree on what happened and who owned the next action.
- What I owned
- Product architecture, data contracts, critical workflows, integrations, production reliability, and delivery practice.
- What changed
- Aura can operate from shared workflow and outcome definitions instead of disconnected sales tools.
Read the story
hyranAgentic orchestration for fashion
A multi-product platform for connected fashion operations
A multi-product software platform connecting brand planning and factory execution through secure contracts, durable workflows, and a shared design foundation.
- The problem
- Brand planning and factory execution needed to connect without collapsing separate users, data, permissions, and release paths into one system.
- What I owned
- Product boundaries, tenant isolation, workflows, migrations, forecasting services, shared contracts, design systems, and release discipline.
- What changed
- Brand and Factory can evolve independently while planning, factory, and model-backed capabilities run behind explicit contracts.
Read the story
What scale changed.
Lessons carried into every Modh engagement.
Scale changes what good engineering requires.
I worked on production systems at the companies below. Aura and Hyran are the work shown here as Modh engagements.
Shopify and Brex
High-growth products demand clear ownership, safe change, and systems that keep working while the company moves quickly.
Motorola Solutions
Operational software has to survive real constraints, long lifecycles, and consequences that extend beyond a release date.
Pfizer and accesso
Enterprise delivery means working through complex stakeholders, integrations, controls, and systems already in motion.
One owner across the hard parts.
Product, platform, and AI decisions rarely stay in separate boxes. I work across all three when the outcome demands it. The engagement ends with working software and a team that can carry it.
01
Product engineering
Turn the real workflow into a product people can depend on.
- 01Customer journeys, product decisions, architecture, and hands-on implementation
- 02Auth, data, integrations, operational tooling, analytics, and revenue workflows
- 03Production releases tied to working user journeys, not feature theatre
02
Software and platform engineering
Build or repair the systems underneath the product.
- 01Platform architecture, APIs, data models, tenant isolation, authorization, and integrations
- 02Modernisation, migrations, observability, performance, and incident repair
- 03Engineering standards and design systems proven through shipped software
03
AI engineering
Put AI into production where it changes the work.
- 01AI product features, agent workflows, tools, evaluations, and guardrails
- 02Provider integration, data access, failure handling, and human review
- 03Team workflows that use AI without lowering the engineering bar
One senior engineer across the product, the platform, and the AI layer. No account team in the middle.
Diagnose it. Build it. Put it in your hands.
I carry the problem through production, then transfer the system, decisions, and operating knowledge to your team.
Diagnose
See reality before touching the code.
Map the workflow, the money path, the failure modes, the data, and the decision that is actually blocked.
Build
Ship the smallest system that changes the outcome.
Protect the data model. Work in vertical slices. Make every release observable, reversible, and useful.
Transfer
Leave the system and the team stronger.
Stabilize the edges, install the operating knowledge, document the calls, and hand over real ownership.
Notes from the work.
01
Your Company Needs Three AI Strategies (You Probably Have Zero)
Every company I talk to has an AI strategy. Product AI. Cool. But that's one out of three. You're missing two-thirds of the opportunity. And your competitors...
02
Stop Building MVPs: Build This Instead
Everyone tells you to build an MVP. Build it cheap. Build it fast. Validate first. And it makes total sense, until six months later you're staring at a...
Bring me the problem that cannot stay stuck.
Email me the product, platform, delivery, or AI problem. Include what is live, what is stuck, and what a useful outcome would look like. I will reply with a direct next step.
You will hear from me, not an account manager.