Senior Technical Specialist, AI Solutions : Delhi

Role Overview

This is a hybrid role at the intersection of solution design and hands-on delivery. You will own enterprise engagements end to end: run technical discovery, design the solution, then build, deploy, and harden production agentic systems inside the customer's environment. You are the person who designs the agent architecture on the whiteboard on Tuesday and debugs its state machine in production on Thursday. You will work directly with customer engineering, operations, and executive teams — translating messy real- world workflows into deterministic, reliable AI agents on the DevRev platform, and proving business impact with numbers, not slideware. This is not an advisory role. You will be measured on working systems in production and the outcomes they

What You Will Do

Discover & Design

Lead technical discovery: map customer workflows, systems of record, and failure modes; identify where agentic automation creates measurable economic impact (cost per case, resolution time, deflection).

Build and deliver PoCs and proofs of value on live customer data and systems — not canned environments.

Translate business problems into agent architectures: orchestration design, state management, tool/skill decomposition, guardrails, and escalation paths.

Shape SOWs, solution designs, and delivery plans with defensible technical scope and effort estimates.

Build & Deploy

Implement production agents on the DevRev platform: agent instructions, deterministic workflows, skills, and integrations with enterprise systems (CRM, ERP, payments, core operational APIs).

Own reliability: design evaluation harnesses, debug agent failures (hallucination, state drift, loops, guardrail leakage), and iterate to production-grade adoption.

Drive integration work hands-on — REST APIs, webhooks, auth flows, data mapping — in collaboration with Product and Engineering.

Instrument and report outcomes: adoption, containment, accuracy, and cost metrics that stand up in a QBR.

Scale & Enable

Convert engagement learnings into reusable assets: reference architectures, playbooks, sandbox environments, and internal tooling.

Enable partner engineers and integrators to deliver independently; run workshops and bootcamps.

Feed structured field signal to Product: what breaks, what's missing, what competitors are doing.

What We're Looking For

Must have

8–10 years in customer-facing technical roles (forward-deployed / field engineering, solution delivery, technical specialist roles) in B2B SaaS or enterprise software.

Hands-on experience building LLM-based systems in production: agent orchestration, prompt and context engineering, RAG, tool/function calling, and evaluation. You have shipped something real, with users, and can walk us through its failure modes.

Strong engineering fundamentals: proficient in Python or JavaScript/TypeScript; fluent with REST APIs, webhooks, and integration patterns; comfortable with SQL and working directly in customer data.

Strong discovery and solutioning craft: stakeholder workshops, PoCs, and technical scoping with clear effort estimates.

Executive-grade communication: you can whiteboard an architecture for engineers and frame ROI for a CXO in the same meeting.

Operates well in ambiguity: unstructured environments, contested stakeholder maps, shifting scope. Bias to ownership.

Strong plus

Experience with deterministic workflow / state-machine design for AI agents, and with making non- deterministic systems reliable enough for regulated or high-volume operations.

Production experience in complex enterprise domains: airlines, BFSI, telecom, manufacturing, or similar multi-system environments.

Familiarity with CRM/CX platforms (Salesforce, Zendesk, ServiceNow), cloud platforms (AWS/Azure/GCP), and modern DevOps tooling.

Exposure to AI-assisted development workflows (Cursor, Copilot, Claude Code).

Logistics

Bachelor's or Master's degree in Computer Science, Engineering, or related field.

This is a customer-embedded role: you will be based on-site at client locations, working face-to-face with customer teams as the default mode of operation — not remote-first with occasional visits. Expect sustained, full-time presence at the customer's offices for the duration of each engagement.