Skip to content
Software Muskets

Fig 7.0AI · Gurgaon · 3–8 yrs · Full-time

AI/ML Engineer

Ship agents, retrieval systems and classical ML models to production, with evals that prove they work and monitoring that proves they still do.

Apply

Email a CV and a few lines about something you built and what you would do differently. No cover letter needed.

Apply by email
Location
Gurgaon
Experience
3–8 yrs
Type
Full-time
Posted
2026-10-24

Fig 7.1The role

What you will be doing.

You will work on our AI Suite (TriageIQ, PolicyDesk, AskCXO) and on client AI projects. The job is production AI: retrieval pipelines, multi-agent orchestration, text-to-SQL with validation, eval harnesses, guardrails, cost controls, and the occasional gradient-boosted model where it beats a language model. You will measure everything and be sceptical of demos, including your own.

  • 01

    Design and build retrieval systems, agents and LLM features with Claude, GPT and open-weight models

  • 02

    Build evaluation sets from real data and automate scoring on every change

  • 03

    Implement guardrails: input and output checks, PII handling, injection defence, refusal logic

  • 04

    Train and deploy classical ML models where they are the right tool, with monitoring and retraining loops

  • 05

    Instrument cost and latency, and make the trade-offs visible

  • 06

    Write up findings for clients and colleagues in plain language

Fig 7.2Must have

What we need to see.

  • 01

    3 to 8 years in software or ML engineering, with at least one LLM-based system in production

  • 02

    Strong Python; comfort with TypeScript is welcome

  • 03

    Hands-on with embeddings, vector and hybrid search, and prompt design as engineering

  • 04

    Experience evaluating model outputs systematically

  • 05

    Solid grounding in statistics and classical ML (scikit-learn, XGBoost, PyTorch)

Fig 7.3Nice to have

Helps, not required.

  • Text-to-SQL or BI semantic layers
  • Speech recognition
  • MLflow, SageMaker or Vertex
  • Hindi or another Indian language

Fig 7.4How we hire

Four steps, two weeks.

  1. 01

    Apply

    Send a CV and a few lines about something you built and what you would do differently. No cover letter needed.

  2. 02

    Intro call

    30 minutes with an engineering lead. We tell you about the work; you tell us what you want.

  3. 03

    Technical round

    A short take-home or a live session, your choice. Real problems from our codebase, AI tools allowed.

  4. 04

    Final conversation

    With leadership. Expectations, growth, compensation. Offer within a week of this call.

Fig 7.5Other roles

Also open.

Apply for AI/ML Engineer.
We reply within a week.

A CV and a few lines about your work are enough. We tell you where you stand after every step.

Apply by email All roles