Founding Forward Deployed Engineer (Agentic Systems)
Posted 23 days ago
The role in short
Deploy autonomous AI agents into enterprise workflows by building agentic systems, implementing integrations, and owning production reliability end-to-end.
About Us
Prentis is an AI research lab building computer-use models that perceive screens and operate software across browser, desktop and mobile. We deploy those models, and the researchers who trained them, into the operations that decide an enterprise's cost, speed and margin.
Position Overview
We’re hiring a Forward Deployed Engineer (FDE) to deploy autonomous AI agents into real enterprise workflows. This is a ship-the-product-in-the-field role: you’ll work directly with customers to map messy processes, build and integrate agentic systems (tools + planning + memory), and own production reliability end-to-end. We’re looking for someone with hands-on experience shipping agentic workflows (tools + planning + state/memory) and the engineering rigor to support them in production (evals + observability).
Role & Responsibilities
Core Responsibilities
- Own 0→1 enterprise deployments: discovery → prototype → production rollout → iterate until measurable ROI
- Design and build agentic systems with tool calling, planning/execution loops, state/memory, and human-in-the-loop (HITL) controls
- Implement reliability mechanisms: retries, timeouts, checkpointing, idempotency, and safe failure modes
- Build enterprise integrations (e.g., Slack/Jira/Salesforce/ServiceNow/SQL/SharePoint/knowledge bases), including auth, permissions, and auditability
- Create evaluation and monitoring systems: offline regression suites, dataset replay, and production quality/cost/latency dashboards
- Debug real failures in production: non-determinism, flaky tools, prompt regressions, partial outages, and “agent got stuck” scenarios
- Collaborate with the founding team to shape product roadmap, technical strategy, and deployment playbooks
Technical Problem Areas (Not Exhaustive)
- Long-running agent workflows with persistent state and memory across sessions
- Orchestration patterns for multi-step, multi-tool processes
- Guardrails, verification, and human approval gates for high-stakes actions
- Reliability + observability (tracing, structured logs, prompt/tool versioning, replay)
- Enterprise security requirements: access control, audit logs, compliance constraints
- Evaluation and benchmarking of agent performance in real production environments
Required Qualifications
- Proof you’ve shipped agentic systems: built and deployed at least one system that includes tool use + planning + state/memory + reliability mechanisms (work or OSS)
- Experience with evaluation of LLM/agent behavior (task suites, regression tests, acceptance metrics, quality thresholds)
- Production engineering mindset: tracing/logging, incident-style debugging, rollout/rollback strategies, versioning prompts/tools
- Strong software engineering fundamentals (Python and/or TypeScript), API integrations, and systems thinking
- Comfortable working directly with customers and owning outcomes end-to-end (Forward Deployed / customer-zero mentality)
- Willingness to work onsite in Manhattan, New York
Preferred Qualifications
- Experience with agent frameworks (e.g., LangGraph, LangChain, LlamaIndex, AutoGen, CrewAI, DSPy) in real deployments
- Workflow engines for long-running processes (e.g., Temporal/Cadence, Airflow/Dagster as applicable)
- Built agent observability: OpenTelemetry-style traces, replay pipelines, prompt/version registries, cost/latency instrumentation
- Experience with enterprise software integration patterns and security/compliance constraints (PII, access control, auditability)
- Familiarity with fine-tuning and adaptation methods (RAG, routing, lightweight fine-tunes, tool augmentation)
Who You Are
- You love deploying real systems into messy environments and iterating fast
- You think in systems: failure modes, tradeoffs, correctness, reliability, scalability
- You’re pragmatic and execution-oriented, but you care deeply about engineering quality
- You’re excited about the limits of agents and motivated to push them responsibly in production
- You can communicate clearly with both technical teams and enterprise stakeholders
What We Offer
- Competitive compensation + meaningful equity, aligned with early-stage risk/reward
- Direct access to the founding team and high ownership from day one
- Resources to ship: strong engineering support and the compute/tools needed to deploy real agentic systems
- A high-velocity environment where you’ll help define the product, the architecture, and the deployment playbook
- Benefits and perks will be finalized as we scale, and we’ll share specifics during the process
Culture & Environment
- Direct access to the founding team and fast, collaborative decision-making
- Opportunity to shape product direction and deployment strategy from day one
- Fast-paced environment with high ownership and high impact
Location
- Onsite in Manhattan, New York with core collaboration hours expected. We believe proximity matters for early-stage execution, especially while defining architecture, product, and deployment playbooks.
Equal Opportunity
- We are committed to building a diverse and inclusive team. We welcome applications from candidates of all backgrounds and are dedicated to fostering an equitable workplace.
Prep for this interview
5 questions likely to come up at Prentis, each matched to a line in this posting.
- 1LLMs, RAG and agents
They wrote: “Familiarity with fine-tuning and adaptation methods (RAG, routing, lightweight fine-tunes, tool augmentation)”
A customer wants a chatbot over their internal documents. How do you build it?
How to answer
Why they ask: RAG is the most common first project. They want the whole path, from messy files to answers people trust.
- Collect and clean the documents; handle permissions per user
- Split them into chunks, embed, and store with source links
- Retrieve, then answer with citations; say “I don't know” when retrieval is weak
- Build a test set of real questions with expected answers before launch
Tip: Spend more time on the data and on testing than on the model.
- 2Integrations and APIs
They wrote: “Strong software engineering fundamentals (Python and/or TypeScript), API integrations, and systems thinking”
Walk me through connecting our product to a customer's internal system that has no documentation.
How to answer
Why they ask: This happens weekly in the job. They want a calm, step-by-step method.
- Find who owns the system and get read access to a test environment
- Watch real traffic or logs to learn the data shapes
- Build a thin adapter with logging around every call
- Agree on how failures get reported, and who fixes what
Tip: Talk about finding the owner of the system first. The code is the easy part.
- 3Cloud, deployment and security
They wrote: “Enterprise security requirements: access control, audit logs, compliance constraints”
The customer's security team sends a 200-question review. How do you handle it?
How to answer
Why they ask: Security reviews often block deals. FDEs help get through them.
- Start from your company's standard answers (SOC 2 report, past reviews)
- Answer clearly; don't oversell
- Bring in your security team for anything you're unsure of
- Keep the new answers for the next customer
Tip: Reuse past answers, flag honest gaps, and offer a call for the hard questions.
- 4Discovery and scoping
They wrote: “Own 0→1 enterprise deployments: discovery → prototype → production rollout → iterate until measurable ROI”
A customer says “we want AI in our claims process.” What do you ask in the first meeting?
How to answer
Why they ask: Most FDE work starts this vague. They want to hear you find the real problem before you pick a tool.
- Map the current process: who touches a claim, in what order, with which systems
- Find the slowest or most error-prone step and ask what it costs them
- Agree on one measurable goal (for example time per claim, or error rate)
- Ask what data you can access, and who signs off on security
Tip: Ask about the work, not the technology. Who does it today, how long it takes, what a mistake costs, and how they'd know it got better.
- 5Data and SQL
They wrote: “Build enterprise integrations (e.g., Slack/Jira/Salesforce/ServiceNow/SQL/SharePoint/knowledge bases), including auth, permissions, and auditability”
Write a query that finds each customer's first and most recent order.
How to answer
Why they ask: A quick check that you're comfortable with grouping, joins and window functions.
- Clarify the tables and what counts as an order (cancelled ones?)
- Use MIN/MAX with GROUP BY, or ROW_NUMBER() if you need the whole row
- Handle ties on the same timestamp, and customers with no orders
- Explain how you'd test it on a few known customers
Tip: Say what you'd check in the data before writing the query: duplicates, time zones, nulls.
Not the one? Get the next one by email
New roles land here every day. Pick what you want to hear about.
Want the job, not just the posting? Learn how FDEs work.
FDE Info is our free weekly newsletter for forward deployed engineers: tools, field notes from real deployments, and how FDE teams hire and pay.
Free. One email a week. Unsubscribe in one click.

