Engineering studio — Bengaluru

We build software
that thinks.

An engineering studio building intelligent, open-source-first systems — from custom apps and ERPs to explainable AI. Deployable anywhere. Owned entirely by you.

"Solo-founded, building toward a team and our own products. One person, strong opinions, real systems in production."

— Founder, CommitX
commitx — zsh
What we believe

Principles, not slogans.

Every architecture decision, every tool choice, every client conversation runs through these filters.

01

Your software should outlive us

Built on open-source foundations — Frappe, SvelteKit, PostgreSQL — so if we're gone tomorrow, your system keeps running. No proprietary lock-in. No hostage negotiations.

02

AI should explain its homework

Every AI system ships with an audit trail. RAG sources are cited. Agent decisions are logged. If the machine can't say why, it doesn't ship.

03

Deploy anywhere, depend on no one

AWS, GCloud, Azure, bare metal — we architect for portability using open formats. S3-compatible storage, not S3. Postgres, not Aurora. Open standards, not proprietary APIs.

04

Track everything, hide nothing

We run our own work on Mattermost, GitHub Projects, and Frappe. Every decision is documented, every status is visible. You get the same transparency we demand of ourselves.

05

Studio now. Products soon.

Building a studio that ships real systems for real companies — and banking the patterns that will become our own products. This is year one of a longer story.

What we build

From first commit to production.

Complete systems — designed, engineered, deployed, and handed over with the keys.

Custom Software & Apps
Core

Full-stack applications designed around your workflows. SvelteKit, React, or Vue frontends. Go or Python backends. Clean APIs.

🧠 AI & Knowledge Systems
AI-native

RAG pipelines, multi-agent orchestration, domain-specific reasoning. Grounded in your data, governed by your rules.

📊 ERP (Frappe / ERPNext)
FOSS

Open-source enterprise resource planning. Accounting, HR, manufacturing — properly integrated, fully yours.

🏥 Healthcare-Grade Systems
Enterprise

EMR, AI-assisted care planning, enterprise scheduling. Built to compliance standards with full audit trails.

🔧 Infrastructure & DevOps
Infra

Kubernetes, dedicated VMs, or managed contracts with SOPs. Self-hosted or cloud — we deploy anywhere and document everything.

🔗 Integration & Migration
Glue

Connect existing systems, migrate off legacy platforms, build bridges between tools that refuse to talk to each other.

AI Systems

We don't build chatbots. We build explainable knowledge systems.

Retrieval-augmented generation, multi-agent orchestration, domain-specific reasoning. Every decision traceable. Every source cited.

📄
Ingest
🧬
Embed
🗄️
Store
🔍
Retrieve
🤖
Reason
🔒
Audit
↑ hover each node — real pipeline architecture
EMR Knowledge Retrieval

Clinical teams query medical knowledge bases with cited, auditable answers. Not ChatGPT — a system that knows your protocols.

Enterprise Search + Citations

SOPs, policies, compliance docs — searchable with semantic understanding and source citations. Replace the shared drive chaos.

AI-Assisted Planning

Project and care planning informed by domain knowledge. The AI drafts, the human decides. Full audit trail on every recommendation.

🚫

What we don't do: "Add AI" as a feature checkbox. No magic wrappers around GPT-4 with your logo. No black boxes. If you want a chatbot widget, there are cheaper options. If you want a knowledge system your team trusts — let's talk.

Open-source stack

If you can't leave, it's not a partnership.

Every tool is open source or uses open formats. You can inspect it, fork it, self-host it, replace us, or run it yourself. No proprietary APIs. No vendor lock-in architectures.

AI & Knowledge
🦙LlamaIndex
🔗LangChain
Qdrant
Weaviate
🎨Chroma
🦙Ollama
Backend
FastAPI
🔵Go
🛠Frappe
🐍Python
💚Node.js
Frontend
🔶SvelteKit
⚛️React
💚Vue
📘TypeScript
🎨Tailwind
Data & Storage
🐘PostgreSQL
Supabase
🔴Redis
🪣S3-compatible
📊ERPNext
Infrastructure
☸️Kubernetes
🐳Docker
🐧Linux
🏗️Terraform
🔀Nginx
🔒Caddy
DevOps & Scaling
🔄GitHub Actions
📈Prometheus
📊Grafana
📝Loki
🤖Ansible
📦Git
🔓

Open formats, not open-washing

S3-compatible object storage, not proprietary blob stores. PostgreSQL, not vendor-locked managed databases. OCI containers, not platform-specific runtimes. We use open standards so you can move between AWS, GCloud, Azure, or bare metal without rewriting a single line.

📦

You own the code

Full repository access from day one. No proprietary layers. Every line is yours.

🗄️

You own the data

Self-hosted or your cloud. Export everything, anytime, standard formats. S3-compatible, always.

🏗️

You own the infra

We set it up, document it, write the SOPs, and hand over the keys. You can run it without us.

Deployment & Infrastructure

Your infrastructure, your rules.

We don't prescribe where your software runs. We architect for portability and deploy wherever makes sense for your organization — with full documentation and SOPs for every scenario.

🏢

In-House / On-Premise

Dedicated VMs or bare metal in your data center. Full air-gap support. You control the hardware, we configure the stack.

Dedicated VMsBare metalAir-gappedFull control
☁️

Cloud (No Lock-In)

AWS, GCloud, Azure — your account, your region. We use open formats (S3-compatible, OCI, Postgres) so you can migrate anytime.

AWSGCloudAzureOpen formatsYour account
📋

Managed + SOPs

We manage the infrastructure under contract with full SOPs, runbooks, and escalation procedures. Transparent, documented, auditable.

SLA contractsRunbooksSOPs24/7 monitoring
Architecture principles
☸️
Kubernetes or simple

K8s for scale and orchestration when needed. Docker Compose + systemd when it's not. We don't over-engineer.

🔄
CI/CD from day one

GitHub Actions pipelines, automated testing, staging environments. Every merge to main is deployable.

📈
Observable by default

Prometheus + Grafana for metrics. Loki for logs. Alerting that pages you before users notice.

🔐
Security as architecture

TLS everywhere. Secrets management. Network policies. RBAC. Not an afterthought — a design constraint.

📦
Reproducible environments

Infrastructure as code via Terraform/Ansible. Every environment is documented and reproducible from scratch.

🌐
DNS, CDN, edge

Proper domain management, CDN configuration, edge caching. Production-grade from the network layer up.

How we work

We track and audit everything. Including ourselves.

Transparent communication isn't a feature we offer — it's how we operate internally. We've adopted these systems for ourselves first, then extended them to every client engagement. You see what we see.

Our internal stack (shared with clients)
💬
Mattermost
Team & client communication
📒
Playbooks
Incident response & checklists
📋
GitHub Projects
Sprint planning & issue tracking
📊
Frappe / ERPNext
Project management & invoicing
📝
Git
Every change tracked, every decision documented
📐
SOPs & Runbooks
Documented procedures for every scenario
What every client gets
Dedicated Mattermost channel — async, threaded, searchable
Live project board — sprints, priorities, blockers visible 24/7
Weekly status updates — no chasing, no surprises
Git access from day one — watch every commit in real-time
Documented SOPs — for deployment, incidents, and handover
Time tracking tied to invoices — know exactly what you're paying for
A typical day — everything is logged
all systems operational sprint 4 · day 3
09:14
sprint.planned
Sprint 4 — 8 stories, 34 points
github-projects
09:42
deploy.staging
v2.3.1-rc1 → staging-east
github-actions
10:15
review.approved
PR #247 — RAG pipeline refactor
github
11:30
client.updated
Weekly status posted to channel
mattermost
14:05
incident.opened
Playbook: elevated response times
playbooks
14:12
incident.resolved
Root cause: connection pool — patched
playbooks
16:00
invoice.generated
Sprint 4 — time tracked, approved
erpnext

"We adopted these tools for ourselves first — Mattermost for comms, GitHub Projects for planning, Frappe for project management and invoicing. What you get isn't a show. It's our actual workflow, extended to include you."

Engagement phases
Phase 01

Understand

We listen, document, and map your domain before touching code.

1–2 wk
Phase 02

Design

Systems architecture, data models, UX flows. All documented, all reviewed.

2–3 wk
Phase 03

Build

Weekly demos, live project boards, daily commits you can watch.

6–16 wk
Phase 04

Ship & Handover

Deployment, SOPs, runbooks, monitoring, knowledge transfer. You get the keys.

1–2 wk

Ready to build something real?

No pitch decks. No discovery calls that go nowhere. Tell us what you're building and we'll tell you how we'd approach it.

hello@commitx.in — Bengaluru, India