Senior
- Ahmedabad or remote within India
- Full-time
gStride AI is hiring a Senior AI Product Engineer to ship productivity intelligence features end-to-end. We are a four-person AI-first product team building productivity intelligence for mid-market teams. We read what teams ship — not how they type.
Apply: email hello@gstride.ai
- book 15 minutes with Ashok
- DM on LinkedIn.
Why this role exists
gStride AI is a four-person product team building AI productivity intelligence for mid-market teams. We read what teams ship, not how they type. The AI Product Engineer pod is two senior engineers who own vertical-slice features end-to-end. Roadmap items — focus mosaic scoring, recommendation engine, audit-trail ML, anomaly detection — are queued behind engineering capacity. That is what this hire fixes.
You join the pod as the third engineer. You become a peer of two senior engineers, paired directly with the AI Product Lead on architecture, and you ship product surface features that customers use within one to three weeks of writing them.
If you have been waiting for a role where AI is the substrate of the work rather than a feature bolted onto a SaaS, this is that role.
What you will do
- Ship end-to-end AI features in 1–3 week vertical slices: data pipeline, model, API, product surface.
- Wire LLM-augmented workflows into the product — focus-mosaic scoring, recommendation generation, anomaly detection.
- Build the audit-trail layer. Every AI score must be reproducible to source events. The explainability and rule-trace stack from Pillar five is your home.
- Own ML eval pipelines: golden-dataset labeling, inter-rater agreement, A/B harnesses, regression gates.
- Work directly with Ashok on product specs. No PM layer.
- Co-design integrations with the four-layer architecture: Capture, Signal, Recommendation, Action.
- Ship a feature every two to three weeks. Cadence is measured in PRs merged, not lines of code.
- Dogfood the product. gStride runs on gStride internally — your dashboards include you.
What we expect
- 4–7 years software engineering experience. Python or TypeScript or Node.js as your primary stack.
- At least one production ML or AI feature you shipped end-to-end — not just notebooks, not just Kaggle.
- Strong fundamentals: SQL, system design, API design, data modeling.
- LLM-fluent. Comfortable with prompt engineering, RAG, agent workflows, and evaluation.
- Comfort with the Azure stack. We run on Azure Static Web Apps and Azure Functions plus Azure AI services for enterprise tenants.
- Anti-surveillance and privacy-first values alignment. No fence-sitters on the positioning.
- Native or near-native English. Bonus: Hindi or Gujarati for local team flow.
- Strong written voice. We ship docs as much as code.
- Comfortable with founder-led ambiguity. We will move the goalposts more than once a quarter.
The stack we run
- Frontend: Vue 3 (static HTML plus Vue components) on Azure Static Web Apps.
- Backend: Node.js Azure Functions today; Python ML services on the roadmap.
- Data: PostgreSQL plus Redis cache. HNSW vector index via AgentDB for semantic search.
- AI: Anthropic Claude as primary, OpenAI as fallback, Azure OpenAI for enterprise tenants.
- Tooling: Claude Code, Cursor, GitHub Copilot — we use all three depending on workflow.
- Observability: GA4, Microsoft Clarity, custom event log.
What you get
- INR 25–50L base salary plus ESOP grant. Final offer indexed to seniority and last comp.
- Ahmedabad office or remote-flexible within India.
- Direct founder access. No middle layer between you and Ashok.
- Real product ownership. Your code ships to all customers.
- AI-augmented every workflow. Claude Code, Cursor, Copilot, and internal tooling on tap.
- Health insurance to Indian standard.
- Conference and learning budget — currently 50K INR per year.
What we do not want
- Anyone who treats AI as wrap-an-LLM-around-it.
- Process-for-process engineers. We do not run scrum theatre.
- Fence-sitters on the anti-surveillance positioning.
- Notebook-only ML. We ship to production, not to Kaggle leaderboards.
- Anyone uncomfortable with directly-typing-with-Claude-Code culture.
How to apply
- Email hello@gstride.ai with subject "AI Product Engineer — your name".
- Include 2–3 sentences on why this role, your LinkedIn URL, and one GitHub link to AI or ML production work you shipped.
- If it clicks, book 15 minutes with Ashok at cal.com/gstrideai. The process from there is two short steps.
Hiring
- Senior
- India
- INR 25–50L + ESOP AI Product Engineer at gStride AI
Ship productivity intelligence features end-to-end — data pipeline, model, API, product surface. Ahmedabad or remote within India. Direct founder access. AI-augmented from day one.
Apply by email Book 15 minDM on LinkedIn
Why this role exists
Engineering is the bottleneck — that is what this hire fixes
gStride AI is a four-person product team building AI productivity intelligence for mid-market teams. We read what teams ship, not how they type. The AI Product Engineer pod is two senior engineers who own vertical-slice features end-to-end. Roadmap items — focus mosaic scoring, recommendation engine, audit-trail ML, anomaly detection — are queued behind engineering capacity. That is what this hire fixes.
You join the pod as the third engineer. You become a peer of two senior engineers, paired directly with the AI Product Lead on architecture, and you ship product surface features that customers use within one to three weeks of writing them.
If you have been waiting for a role where AI is the substrate of the work rather than a feature bolted onto a SaaS, this is that role.
What you will do
The job, in eight lines
Ship end-to-end AI features
1–3 week vertical slices. Data pipeline, model, API, product surface. You own the whole vertical.
Wire LLM-augmented workflows
Focus-mosaic scoring, recommendation generation, anomaly detection. Prompt + RAG + eval all in your hands.
Build the audit-trail layer
Every AI score must be reproducible to source events. The explainability and rule-trace stack from Pillar five is your home.
Own ML eval pipelines
Golden-dataset labeling, inter-rater agreement, A/B harnesses, regression gates before deploy.
Co-spec with the founder
No PM layer. You sit with Ashok, sharpen the spec, and ship. Decisions in days, not quarters.
Co-design the 4-layer architecture
Capture, Signal, Recommendation, Action. You wire integrations across all four layers.
Ship a feature every 2–3 weeks
Cadence is measured in PRs merged and eval scores, not lines of code or hours logged.
Dogfood gStride internally
gStride runs on gStride. Your own dashboards include your work. Bugs you cause hit you first.
Ship end-to-end AI features
1–3 week vertical slices. Data pipeline, model, API, product surface. You own the whole vertical.
Wire LLM-augmented workflows
Focus-mosaic scoring, recommendation generation, anomaly detection. Prompt + RAG + eval all in your hands.
Build the audit-trail layer
Every AI score must be reproducible to source events. The explainability and rule-trace stack from Pillar five is your home.
Own ML eval pipelines
Golden-dataset labeling, inter-rater agreement, A/B harnesses, regression gates before deploy.
Co-spec with the founder
No PM layer. You sit with Ashok, sharpen the spec, and ship. Decisions in days, not quarters.
Co-design the 4-layer architecture
Capture, Signal, Recommendation, Action. You wire integrations across all four layers.
Ship a feature every 2–3 weeks
Cadence is measured in PRs merged and eval scores, not lines of code or hours logged.
Dogfood gStride internally
gStride runs on gStride. Your own dashboards include your work. Bugs you cause hit you first.
Sounds like the role you have been waiting for?
Apply by email Book 15 min
What we expect
Qualifications
4–7 years engineering
Python or TypeScript or Node.js as your primary stack.
Production AI/ML shipped
At least one ML or AI feature you took to production end-to-end — not just notebooks, not just Kaggle.
Strong fundamentals
SQL, system design, API design, data modeling. The boring layer that everything sits on.
LLM-fluent
Prompt engineering, RAG, agent workflows, evaluation. You have shipped, not just read about, all four.
Azure-comfortable
Static Web Apps, Functions, AI services. The stack we run today and the stack we are scaling on.
Anti-surveillance values
Privacy-first alignment. No fence-sitters on the positioning — we read what teams ship, not how they type.
Strong English
Native or near-native. Bonus: Hindi or Gujarati for local team flow.
Strong written voice
We ship docs as much as code. Specs, ADRs, RFCs — written before merged.
Founder-led ambiguity
We will move the goalposts more than once a quarter. You stay calm and ship.
4–7 years engineering
Python or TypeScript or Node.js as your primary stack.
Production AI/ML shipped
At least one ML or AI feature you took to production end-to-end — not just notebooks, not just Kaggle.
Strong fundamentals
SQL, system design, API design, data modeling. The boring layer that everything sits on.
LLM-fluent
Prompt engineering, RAG, agent workflows, evaluation. You have shipped, not just read about, all four.
Azure-comfortable
Static Web Apps, Functions, AI services. The stack we run today and the stack we are scaling on.
Anti-surveillance values
Privacy-first alignment. No fence-sitters on the positioning — we read what teams ship, not how they type.
Strong English
Native or near-native. Bonus: Hindi or Gujarati for local team flow.
Strong written voice
We ship docs as much as code. Specs, ADRs, RFCs — written before merged.
Founder-led ambiguity
We will move the goalposts more than once a quarter. You stay calm and ship.
The stack
What we run in production
Frontend
Vue 3 — static HTML + Vue components on Azure Static Web Apps.
Backend
Node.js Azure Functions today; Python ML services on the roadmap.
Data
PostgreSQL + Redis cache. HNSW vector index via AgentDB for semantic search.
AI
Anthropic Claude (primary), OpenAI (fallback), Azure OpenAI (enterprise tenants).
Tooling
Claude Code + Cursor + GitHub Copilot — all three, depending on the workflow.
Observability
GA4 + Microsoft Clarity + custom event log.
Frontend
Vue 3 — static HTML + Vue components on Azure Static Web Apps.
Backend
Node.js Azure Functions today; Python ML services on the roadmap.
Data
PostgreSQL + Redis cache. HNSW vector index via AgentDB for semantic search.
AI
Anthropic Claude (primary), OpenAI (fallback), Azure OpenAI (enterprise tenants).
Tooling
Claude Code + Cursor + GitHub Copilot — all three, depending on the workflow.
Observability
GA4 + Microsoft Clarity + custom event log.
What you get
Compensation and working setup
INR 25–50L base + ESOP
Final offer indexed to seniority and last comp. ESOP grant on top.
Ahmedabad or remote
Office in Ahmedabad. Remote-flexible within India for the right candidate.
Direct founder access
No middle layer between you and Ashok. Specs and decisions happen in the same room.
Real product ownership
Your code ships to all customers within one to three weeks of writing it.
AI-augmented workflows
Claude Code, Cursor, Copilot, internal tooling on tap. Paired with AI by default, not as an afterthought.
Health insurance
To Indian-standard cover. Standard practice — listed so it is on the page.
Learning budget
Currently 50K INR per year for conferences, courses, books. Use it.
INR 25–50L base + ESOP
Final offer indexed to seniority and last comp. ESOP grant on top.
Ahmedabad or remote
Office in Ahmedabad. Remote-flexible within India for the right candidate.
Direct founder access
No middle layer between you and Ashok. Specs and decisions happen in the same room.
Real product ownership
Your code ships to all customers within one to three weeks of writing it.
AI-augmented workflows
Claude Code, Cursor, Copilot, internal tooling on tap. Paired with AI by default, not as an afterthought.
Health insurance
To Indian-standard cover. Standard practice — listed so it is on the page.
Learning budget
Currently 50K INR per year for conferences, courses, books. Use it.
Anti-pattern flags
What we do not want
Anyone who treats AI as wrap-an-LLM-around-it.
Process-for-process engineers. We do not run scrum theatre.
Fence-sitters on the anti-surveillance positioning.
Notebook-only ML. We ship to production, not to Kaggle leaderboards.
Anyone uncomfortable with directly-typing-with-Claude-Code culture.
Anyone who treats AI as wrap-an-LLM-around-it.
Process-for-process engineers. We do not run scrum theatre.
Fence-sitters on the anti-surveillance positioning.
Notebook-only ML. We ship to production, not to Kaggle leaderboards.
Anyone uncomfortable with directly-typing-with-Claude-Code culture.
How to apply
- Email hello@gstride.ai with subject "AI Product Engineer — your name".
- Include 2–3 sentences on why this role, your LinkedIn URL, and one GitHub link to AI or ML production work you shipped.
- If it clicks, book 15 minutes with Ashok at cal.com/gstrideai. The process from there is two short steps.
Apply by email Book 15 min