---
schema: 1
type: video
title: How Forward Deployed Engineering is done at Decagon — Sunny Rekhi
description: >-
  Decagon builds the customer support agent that answers when you email a brand,
  and making that agent good for a specific enterprise is the forward deployed
  job. It splits in two: one side configures the agent brain, the instructions
  and the handoff rules for when a human should take over; the other works like
  product engineering, figuring out what a new enterprise needs and anticipating
  the requests that have not been made yet. Sunny Rekhi's point is that the line
  between forward deployed and product is thin, because a customer ask is often
  just a product feature waiting to be built. What changes at scale, from 50
  people to 500, is that this splits into specialized lanes and a few
  disciplines start to matter. One is restraint: some problems belong to the
  customer, and getting the boundary in writing keeps everyone honest. The
  bigger one is that custom work does not stay custom. Every integration a
  forward deployed engineer builds gets upstreamed into the platform, so custom
  becomes self serve and the agent compounds, and the next customer inherits it
  for free. You still have to prove value fast, because the relationship is
  multi year but the trust is won in the first weeks. Speaker info: -
  https://www.linkedin.com/in/sunny-rekhi/ Timestamps: 0:00 - Introduction:
  forward deployed at Decagon 1:06 - What Decagon's support agent does 2:24 -
  Offloading support and proactive outreach 3:40 - Two kinds of forward deployed
  work 6:12 - Scaling from 50 to 500 people 8:20 - Restraint, and getting it in
  writing 9:59 - Ramping a new enterprise deal 11:18 - Solving a problem so it
  scales 13:00 - Proving value fast 14:18 - Custom becomes self serve
canonical: 'https://liquidgold.cc/watch/7wu2hsRfvV0'
markdown: 'https://liquidgold.cc/watch/7wu2hsRfvV0.md'
llms: 'https://liquidgold.cc/watch/7wu2hsRfvV0/llms.txt'
parent: 'https://liquidgold.cc/channel/aidotengineer.md'
channel: aidotengineer
updated: '2026-10-24T06:30:03.078Z'
language: null
children: []
videoId: 7wu2hsRfvV0
durationSeconds: 1089
publishedAt: '2026-07-28T17:00:35.000Z'
searchChunks:
  - ordinal: 0
    text: null
    startSeconds: 79
    endSeconds: 145
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=79'
    heldReason: rights
  - ordinal: 1
    text: null
    startSeconds: 79
    endSeconds: 145
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=79'
    heldReason: rights
  - ordinal: 2
    text: null
    startSeconds: 145
    endSeconds: 156
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=145'
    heldReason: rights
  - ordinal: 3
    text: null
    startSeconds: 156
    endSeconds: 231
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=156'
    heldReason: rights
  - ordinal: 4
    text: null
    startSeconds: 241
    endSeconds: 374
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=241'
    heldReason: rights
  - ordinal: 5
    text: null
    startSeconds: 241
    endSeconds: 374
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=241'
    heldReason: rights
  - ordinal: 6
    text: null
    startSeconds: 241
    endSeconds: 374
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=241'
    heldReason: rights
  - ordinal: 7
    text: null
    startSeconds: 384
    endSeconds: 506
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=384'
    heldReason: rights
  - ordinal: 8
    text: null
    startSeconds: 384
    endSeconds: 506
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=384'
    heldReason: rights
  - ordinal: 9
    text: null
    startSeconds: 518
    endSeconds: 609
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=518'
    heldReason: rights
  - ordinal: 10
    text: null
    startSeconds: 619
    endSeconds: 683
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=619'
    heldReason: rights
  - ordinal: 11
    text: null
    startSeconds: 694
    endSeconds: 780
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=694'
    heldReason: rights
  - ordinal: 12
    text: null
    startSeconds: 780
    endSeconds: 791
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=780'
    heldReason: rights
  - ordinal: 13
    text: null
    startSeconds: 866
    endSeconds: 878
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=866'
    heldReason: rights
  - ordinal: 14
    text: null
    startSeconds: 878
    endSeconds: 1089
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=878'
    heldReason: rights
  - ordinal: 15
    text: null
    startSeconds: 878
    endSeconds: 1089
    watch: 'https://liquidgold.cc/watch/7wu2hsRfvV0?t=878'
    heldReason: rights
---


# How Forward Deployed Engineering is done at Decagon — Sunny Rekhi

Decagon builds the customer support agent that answers when you email a brand, and making that agent good for a specific enterprise is the forward deployed job. It splits in two: one side configures the agent brain, the instructions and the handoff rules for when a human should take over; the other works like product engineering, figuring out what a new enterprise needs and anticipating the requests that have not been made yet. Sunny Rekhi's point is that the line between forward deployed and product is thin, because a customer ask is often just a product feature waiting to be built. What changes at scale, from 50 people to 500, is that this splits into specialized lanes and a few disciplines start to matter. One is restraint: some problems belong to the customer, and getting the boundary in writing keeps everyone honest. The bigger one is that custom work does not stay custom. Every integration a forward deployed engineer builds gets upstreamed into the platform, so custom becomes self serve and the agent compounds, and the next customer inherits it for free. You still have to prove value fast, because the relationship is multi year but the trust is won in the first weeks. Speaker info: - https://www.linkedin.com/in/sunny-rekhi/ Timestamps: 0:00 - Introduction: forward deployed at Decagon 1:06 - What Decagon's support agent does 2:24 - Offloading support and proactive outreach 3:40 - Two kinds of forward deployed work 6:12 - Scaling from 50 to 500 people 8:20 - Restraint, and getting it in writing 9:59 - Ramping a new enterprise deal 11:18 - Solving a problem so it scales 13:00 - Proving value fast 14:18 - Custom becomes self serve

Canonical: https://liquidgold.cc/watch/7wu2hsRfvV0 · Channel: [AI Engineer](https://liquidgold.cc/channel/aidotengineer.md)

> Search chunk text is held until public-source provenance is verified.

## Document chunks

> Missing: no stored document chunks are available.