Senior AI Engineer / Forward-Deployed Engineer Mumbai, hybrid, client sites 6+ years

Envisstack; Mumbai; Hybrid; INR 4,000,000 - 6,000,000; Full time

Job summary

Envisstack in Mumbai is seeking a senior software engineer to lead AI engagements inside client environments. You will run workshops, define use cases and evaluation plans, design system architecture, and implement core components while guiding junior engineers.

You will also manage security reviews, integrate with enterprise systems, and help shape how we build, measure, and hire as part of a growing team.

Qualifications

• 6+ years building and running software in production, including time as a tech lead or senior engineer.
• ] ,
• job_responsibilities_description=[

Responsibilities

• Lead Discover workshops with operations, IT and leadership teams.
• Design the architecture: model choice, retrieval, agent and tool design, guardrails, hosting and cost.
• Build the hard parts yourself, and review the code of AI Engineers and Interns.
• Own the evaluation: agree the evaluation set and target with the client, and report scores honestly.
• Work through security reviews, data-protection questions and sector rules with client IT, risk and compliance teams.
• Integrate with what the client runs: core banking, SAP, Tally, Salesforce, Zoho, WhatsApp, telephony and legacy databases.
• Plan staged launches, rollback, monitoring and on-call. Hand over cleanly into the Managed AI retainer.
• Write the documents that matter: design notes, risk registers, evaluation reports and runbooks.
• Help us hire, interview and mentor.

Skills

Python

Backend APIs

LLM integration

Architectural design

Retrieval search

Agents and tools

Security & compliance

Client leadership

Mentoring

Deployment

Tools

Docker

CI/CD

Job description

Type Full time

Location Mumbai, hybrid, with regular time at client sites across India

A forward-deployed engineer is the person who turns a client’s problem into a working system inside the client’s walls. You will be the technical lead on our engagements. You will run the workshops, find the part of the problem that matters, design the system, write a lot of the code, and stand in front of the client’s leadership to explain the evaluation results, good or bad.

This is one of the first senior engineering hires at a new firm. You will shape how we build, how we measure and how we hire.

• Lead Discover workshops with operations, IT and leadership teams. Turn what you see into a scoped use case, a success metric and an evaluation plan.
• Design the architecture: model choice, retrieval, agent and tool design, guardrails, integration, hosting and cost.
• Build the hard parts yourself, and review the code of AI Engineers and Interns.
• Own the evaluation: agree the evaluation set and target with the client, and report scores honestly.
• Work through security reviews, data-protection questions and sector rules (DPDP, RBI, IRDAI, SEBI) with client IT, risk and compliance teams.
• Integrate with what the client runs: core banking, SAP, Tally, Salesforce, Zoho, WhatsApp, telephony and legacy databases.
• Plan staged launches, rollback, monitoring and on-call. Hand over cleanly into the Managed AI retainer.
• Write the documents that matter: design notes, risk registers, evaluation reports and runbooks.
• Help us hire, interview and mentor.

You should have

• 6 or more years building and running software in production, including time as a tech lead or senior engineer.
• Shipped at least one LLM-based system to real users, and can explain how you evaluated it and what broke.
• Strong Python, and solid backend skills: APIs, databases, queues, testing, CI/CD and containers.
• Depth in at least two of: retrieval and search, agents and tool use, speech, document AI, evaluation and observability, ML platform and inference.
• Experience integrating with enterprise systems and working inside a client’s security and change processes.
• Calm, clear communication with non-technical leaders. You can say “this will not work, and here is why” in a way people thank you for.
• Willingness to travel to client sites, when the work needs it.

Good to have

• Consulting, solutions engineering or professional services experience.
• Work in Indian BFSI, healthcare, retail or manufacturing.
• One or more Indian languages, spoken and written.
• Experience hosting open-weight models: vLLM or similar, quantisation and GPU cost control.
• Familiarity with the DPDP Act 2023 and its Rules, ISO/IEC 27001 or ISO/IEC 42001.

Client outcomes

Did the system meet the metric agreed in writing, and does it keep meeting it in Run?

Is the system secure, observable, documented and supportable by someone else?