Website:
workfabric.com
Job details:
Workfabric AI is the enterprise simulation company. We build digital twins of every account, team, product, and persona — so the enterprise can simulate decisions before shipping them.
Our product, ContextFabric, is the simulation layer that senses across your apps, synthesizes operational context, and grounds every twin in how your teams actually work. We believe operational context is the real competitive advantage in enterprise AI, and we're building the platform that turns it into decisions.
About the role
We simulate the enterprise by creating digital twins — of people, and of entities such as accounts, teams, and products. This role is about building them. You'll work on building one of these twins, and own problem statements within it — shaping the problem, designing the architecture, writing the code, and shipping it to production. You'll be accountable for the systems that ship, not just the pull request.
What you'll be doing
Three hard problems sit at the centre of this role:
- Ground every twin in real work. Make each twin reason from the enterprise's actual behavioural context, so its output stays grounded and defensible — per tenant, never crossing a boundary.
- Run thousands of twins in parallel. Build the runtime that executes twins at scale — thousands running concurrently and collaborating with each other to do the work.
- Give twins the right tools and permissions. Ensure every twin has the tool sit needs to do its work and the permissions to use them safely — governed by identity, policy, and an audit trail you could replay for a regulator.
The surface area you'll own, and set the patterns for
- The twin runtime that executes and composes twins at scale.
- State management for execution history, checkpointing, and recovery across long-running twin runs.
- Event streaming and messaging for real-time progress, with logging, metrics, and tracing instrumented throughout.
- Authentication and identity across the platform — including access controls and approval workflows for twin operations, in partnership with security.
- The engineering patterns and quality bar for your area — through code review and pairing with the engineers around you.
- The operational backbone the platform runs on — deployment, CI/CD, autoscaling, and reliability as infrastructure-as-code.
Who you are
You may be a good fit if you have
- B.E. / B. Tech from a top university in India in one of the following departments (i) Computer Science, (ii) Information Sciences, (iii) Information Technology (iv) AI/ML
- 5+ years building production software in Python, with strong backend fundamentals in API design and data modelling.
- Hands-on experience building services with FastAPI as your primary framework.
- Hands-on experience building LLM-backed applications in production —orchestrating multi-step reasoning, retrieval, and state — with any orchestration framework (LangChain, LangGraph, or equivalent).
- Comfort designing for concurrency and latency — async, parallelism, and caching — so multi-step systems stay fast under load.
- A track record of owning features or services end to end, from design through production.
- Clear communication and the ability to work autonomously in an ambiguous, fast-moving space.
Strong candidates may also have
- Depth with a specific orchestration framework such as LangChain or LangGraph, plus hands-on RAG and vector databases.
- Experience building evaluations for LLM or agentic systems — measuring quality, catching regressions, and guarding against unsafe behaviour.
- Experience with React for building customer-facing UIs and dashboards.
- Comfortable deploying, operating, and maintaining your own services in a production cloud environment (Azure a plus).
- Infrastructure-as-code with Terraform, plus Kubernetes and Helm for deployment.
- Observability — structured logging, distributed tracing, and metrics.
Click on Apply to know more.
