diff --git a/.gitignore b/.gitignore
index b6e4761..b282e51 100644
--- a/.gitignore
+++ b/.gitignore
@@ -2,6 +2,7 @@
 __pycache__/
 *.py[cod]
 *$py.class
+*.sw[a-z]
 
 # C extensions
 *.so
diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000..64785ab
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,21 @@
+appdirs==1.4.4
+attr==0.3.1
+attrs==20.2.0
+black==20.8b1
+click==7.1.2
+iniconfig==1.0.1
+invoke==1.4.1
+more-itertools==8.5.0
+mypy-extensions==0.4.3
+packaging==20.4
+pathspec==0.8.0
+pluggy==0.13.1
+py==1.9.0
+pyinotify==0.9.6
+pyparsing==2.4.7
+pytest==6.0.2
+regex==2020.7.14
+six==1.15.0
+toml==0.10.1
+typed-ast==1.4.1
+typing-extensions==3.7.4.3
diff --git a/script/bootstrap b/script/bootstrap
new file mode 100644
index 0000000..e644547
--- /dev/null
+++ b/script/bootstrap
@@ -0,0 +1,15 @@
+VENV=codesearch.venv
+
+#################################################################
+# Bootstrapping sets up the Python 3.8 venv that allows the use #
+# of the invoke commands.                                       #
+#################################################################
+
+{
+    pyenv virtualenv-delete -f $VENV
+    pyenv virtualenv $VENV &&
+    pyenv activate $VENV &&
+    python -m pip install -U pip &&
+    pip install -r requirements.txt &&
+    echo "✨ Good to go! ✨"
+}
diff --git a/src/__init__.py b/src/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/src/__snapshots__/test_prefix_tree.ambr b/src/__snapshots__/test_prefix_tree.ambr
new file mode 100644
index 0000000..1ddf955
--- /dev/null
+++ b/src/__snapshots__/test_prefix_tree.ambr
@@ -0,0 +1,184 @@
+# name: test_base_tree_has_a_root_node
+  <class 'dict'> {
+    'children': <class 'list'> [
+    ],
+    'mappings': <class 'list'> [
+    ],
+    'value': None,
+  }
+---
+# name: test_insert_multiple_keys_same_string
+  <class 'dict'> {
+    'children': <class 'list'> [
+      <class 'dict'> {
+        'children': <class 'list'> [
+          <class 'dict'> {
+            'children': <class 'list'> [
+              <class 'dict'> {
+                'children': <class 'list'> [
+                  <class 'dict'> {
+                    'children': <class 'list'> [
+                    ],
+                    'mappings': <class 'list'> [
+                      'key_1',
+                      'key_2',
+                    ],
+                    'value': 'd',
+                  },
+                ],
+                'mappings': <class 'list'> [
+                ],
+                'value': 'c',
+              },
+            ],
+            'mappings': <class 'list'> [
+            ],
+            'value': 'b',
+          },
+        ],
+        'mappings': <class 'list'> [
+        ],
+        'value': 'a',
+      },
+    ],
+    'mappings': <class 'list'> [
+    ],
+    'value': None,
+  }
+---
+# name: test_insert_overlapping_strings
+  <class 'dict'> {
+    'children': <class 'list'> [
+      <class 'dict'> {
+        'children': <class 'list'> [
+          <class 'dict'> {
+            'children': <class 'list'> [
+              <class 'dict'> {
+                'children': <class 'list'> [
+                  <class 'dict'> {
+                    'children': <class 'list'> [
+                    ],
+                    'mappings': <class 'list'> [
+                      'key_1',
+                    ],
+                    'value': 'd',
+                  },
+                  <class 'dict'> {
+                    'children': <class 'list'> [
+                    ],
+                    'mappings': <class 'list'> [
+                      'key_2',
+                    ],
+                    'value': 'e',
+                  },
+                ],
+                'mappings': <class 'list'> [
+                ],
+                'value': 'c',
+              },
+            ],
+            'mappings': <class 'list'> [
+            ],
+            'value': 'b',
+          },
+        ],
+        'mappings': <class 'list'> [
+        ],
+        'value': 'a',
+      },
+    ],
+    'mappings': <class 'list'> [
+    ],
+    'value': None,
+  }
+---
+# name: test_insert_single_character_
+  <class 'dict'> {
+    'children': <class 'list'> [
+      <class 'dict'> {
+        'children': <class 'list'> [
+        ],
+        'mappings': <class 'list'> [
+          'key_1',
+        ],
+        'value': 'a',
+      },
+    ],
+    'mappings': <class 'list'> [
+    ],
+    'value': None,
+  }
+---
+# name: test_insert_single_string
+  <class 'dict'> {
+    'children': <class 'list'> [
+      <class 'dict'> {
+        'children': <class 'list'> [
+          <class 'dict'> {
+            'children': <class 'list'> [
+              <class 'dict'> {
+                'children': <class 'list'> [
+                ],
+                'mappings': <class 'list'> [
+                  'key_1',
+                ],
+                'value': 'c',
+              },
+            ],
+            'mappings': <class 'list'> [
+            ],
+            'value': 'b',
+          },
+        ],
+        'mappings': <class 'list'> [
+        ],
+        'value': 'a',
+      },
+    ],
+    'mappings': <class 'list'> [
+    ],
+    'value': None,
+  }
+---
+# name: test_insert_strings_subsets_of_each_other
+  <class 'dict'> {
+    'children': <class 'list'> [
+      <class 'dict'> {
+        'children': <class 'list'> [
+          <class 'dict'> {
+            'children': <class 'list'> [
+              <class 'dict'> {
+                'children': <class 'list'> [
+                  <class 'dict'> {
+                    'children': <class 'list'> [
+                    ],
+                    'mappings': <class 'list'> [
+                      'key_1',
+                    ],
+                    'value': 'd',
+                  },
+                ],
+                'mappings': <class 'list'> [
+                  'key_2',
+                ],
+                'value': 'c',
+              },
+            ],
+            'mappings': <class 'list'> [
+            ],
+            'value': 'b',
+          },
+        ],
+        'mappings': <class 'list'> [
+        ],
+        'value': 'a',
+      },
+    ],
+    'mappings': <class 'list'> [
+    ],
+    'value': None,
+  }
+---
+# name: test_serializes_to_json
+  '{"value": null, "mappings": [], "children": [{"value": "a", "mappings": [], "children": [{"value": "b", "mappings": [], "children": [{"value": "c", "mappings": [], "children": [{"value": "d", "mappings": ["key_1"], "children": []}]}]}]}]}'
+---
diff --git a/src/codesearch/__init__.py b/src/codesearch/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/src/codesearch/base.py b/src/codesearch/base.py
new file mode 100644
index 0000000..7fc2d3c
--- /dev/null
+++ b/src/codesearch/base.py
@@ -0,0 +1,18 @@
+from abc import ABC, abstractmethod
+from typing import List, Optional
+
+
+class IndexBase(ABC):
+    @abstractmethod
+    def index(self, content: str, haystack: Optional[List[str]]):
+        pass
+
+    @abstractmethod
+    def query(self, query: str) -> List[str]:
+        pass
+
+
+class IndexerBase(ABC):
+    @abstractmethod
+    def index(self, paths: List[str]):
+        pass
diff --git a/src/codesearch/cli.py b/src/codesearch/cli.py
new file mode 100644
index 0000000..5f66ddb
--- /dev/null
+++ b/src/codesearch/cli.py
@@ -0,0 +1,30 @@
+import argparse
+
+from pathlib import Path
+from .server import Server
+from .indexer import Indexer
+from .client import search
+from .settings import settings
+
+
+def main():
+    parser = argparse.ArgumentParser()
+
+    parser.add_argument("command")
+    parser.add_argument("--q", required=False)
+
+    args = parser.parse_args()
+
+    if args.command == "start":
+        watched = [Path(p).expanduser() for p in settings.WATCHED]
+        server = Server(
+            indexer=Indexer(
+                domain=watched,
+                exclusions=settings.EXCLUDES,
+                file_types=settings.FILE_TYPES,
+            ),
+            watched=watched,
+        )
+        server.run()
+    elif args.command == "search":
+        search(args.q)
diff --git a/src/codesearch/client.py b/src/codesearch/client.py
new file mode 100644
index 0000000..e5291e4
--- /dev/null
+++ b/src/codesearch/client.py
@@ -0,0 +1,83 @@
+import socket
+import json
+from pathlib import Path
+
+import curses
+
+
+from .settings import settings
+from .colors import highlight
+
+
+def display_handler(stdscr, buffer):
+    current_y = 0
+    stdscr.refresh()
+    curses.start_color()
+    y, x = stdscr.getmaxyx()
+    curses.init_pair(1, curses.COLOR_GREEN, curses.COLOR_BLACK)
+    pad = curses.newpad(y, x)
+    while True:
+        row = 0
+        y_offset = 0
+        pad.clear()
+        while row < current_y + y - 1:
+            l = buffer[current_y + y_offset]
+            if l["type"] == "path":
+                pad.addstr(row, 0, l["value"], curses.color_pair(1))
+                row += 1
+                y_offset += 1
+            elif l["type"] == "sep":
+                row += 1
+                y_offset += 1
+            else:
+                pad.addstr(row, 0, str(l["lineno"]), curses.color_pair(1))
+                pad.addstr(row, 5, l["value"])
+                row += 1
+                y_offset += 1
+
+            if y_offset == y or current_y == y - 1:
+                break
+
+        pad.refresh(0, 0, 0, 0, y, x)
+        key = stdscr.getch()
+
+        if key in [81, 113]:
+            break
+        elif key == curses.KEY_UP:
+            current_y = max(0, current_y - 1)
+        elif key == curses.KEY_DOWN:
+            current_y = min(len(buffer), current_y + 1)
+        elif key == curses.KEY_NPAGE:
+            current_y = min(len(buffer), current_y + y + 1)
+        elif key == curses.KEY_PPAGE:
+            current_y = max(0, current_y - y - 1)
+
+
+def search(query):
+    with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s:
+        s.connect((settings.SOCKET_HOST, settings.SOCKET_PORT))
+        s.sendall(query.encode())
+        length = int(s.recv(8).decode())
+        results = None
+
+        with open(Path(settings.BUFFER_PATH).expanduser(), "rb") as infile:
+            results = infile.read().decode()
+
+        results = json.loads(results)
+
+        output = []
+        for result in results:
+            with open(result["key"], "r") as infile:
+                highlighted_text = infile.read()[
+                    result["offset_start"] : result["offset_end"]
+                ]
+                line_number = result["line_start"]
+                output.append({"value": result["key"], "type": "path"})
+                for l in highlighted_text.split("\n"):
+                    output.append({"value": l, "type": "code", "lineno": line_number})
+                    line_number += 1
+                output.append({"type": "sep"})
+
+        s.close()
+
+        curses.wrapper(display_handler, output)
diff --git a/src/codesearch/colors.py b/src/codesearch/colors.py
new file mode 100644
index 0000000..98a10d6
--- /dev/null
+++ b/src/codesearch/colors.py
@@ -0,0 +1,7 @@
+COLORS = {"green": "\033[92m", "yellow": "\033[93m", "red": "\033[91m"}
+ENDC = "\033[0m"
+
+
+def highlight(text, color="green"):
+    color_code = COLORS[color]
+    return f"{color_code}{text}{ENDC}"
diff --git a/src/codesearch/constants.py b/src/codesearch/constants.py
new file mode 100644
index 0000000..9a14ce5
--- /dev/null
+++ b/src/codesearch/constants.py
@@ -0,0 +1,11 @@
+SETTINGS_KEYS = [
+    "WATCHED",
+    "SOCKET_PORT",
+    "SOCKET_HOST",
+    "EXCLUDES",
+    "FILE_TYPES",
+    "SIGNIFICANCE_THRESHOLD",
+    "INDEXING_PROCESSES",
+    "BUFFER_PATH",
+]
+QUERY_STRING_LENGTH = 1024
diff --git a/src/codesearch/document_models.py b/src/codesearch/document_models.py
new file mode 100644
index 0000000..1652a33
--- /dev/null
+++ b/src/codesearch/document_models.py
@@ -0,0 +1,52 @@
+import attr
+
+
+@attr.s
+class Corpus:
+    _documents = attr.ib(default=attr.Factory(dict))
+    _key_to_uid = attr.ib(default=attr.Factory(dict))
+
+    @property
+    def document_count(self):
+        return len(self._documents)
+
+    def add_document(self, key, content):
+        document_uid = f"document:{self.document_count}"
+
+        self._documents[document_uid] = Document(
+            uid=document_uid, key=key, content=content
+        )
+        self._key_to_uid[key] = document_uid
+
+        return document_uid
+
+    def get_document(self, uid=None, key=None):
+        if key:
+            uid = self._key_to_uid[key]
+
+        return self._documents[uid]
+
+    def collect_unprocessed_documents(self):
+        return [
+            uid
+            for uid in self._documents
+            if not self.get_document(uid=uid).is_processed
+        ]
+
+    def mark_document_as_processed(self, uid):
+        self._documents[uid].mark_as_processed()
+
+
+@attr.s
+class Document:
+    uid = attr.ib()
+    key = attr.ib()
+    content = attr.ib()
+    _processed = attr.ib(default=False)
+
+    @property
+    def is_processed(self):
+        return self._processed
+
+    def mark_as_processed(self):
+        self._processed = True
diff --git a/src/codesearch/indexer.py b/src/codesearch/indexer.py
new file mode 100644
index 0000000..c4090fa
--- /dev/null
+++ b/src/codesearch/indexer.py
@@ -0,0 +1,211 @@
+from .base import IndexerBase
+from pathlib import Path
+from typing import Dict, List
+import re
+from time import perf_counter
+from multiprocessing import Pool
+import mmap
+
+import attr
+
+from .settings import settings
+
+from .process_utils import chunkify_content
+from .document_models import Corpus
+from .trigram_index import TrigramIndex
+from .line_index import LineIndex
+from .logger import get_logger
+
+logger = get_logger(__name__)
+
+
+@attr.s
+class SearchResult:
+    key = attr.ib()
+    offset_start = attr.ib()
+    offset_end = attr.ib()
+    line_start = attr.ib()
+    line_end = attr.ib()
+
+    def to_dict(self):
+        return {
+            "key": self.key,
+            "offset_start": self.offset_start,
+            "offset_end": self.offset_end,
+            "line_start": self.line_start,
+            "line_end": self.line_end,
+        }
+
+
+@attr.s
+class Indexer(IndexerBase):
+    # Indices
+    _trigram_index = attr.ib(default=attr.Factory(TrigramIndex))
+    _line_index = attr.ib(default=attr.Factory(LineIndex))
+
+    _exclusions = attr.ib(default=attr.Factory(list))
+    _file_types = attr.ib(default=attr.Factory(list))
+    # Document corpus
+    corpus = attr.ib(default=attr.Factory(Corpus))
+    domain = attr.ib(default=attr.Factory(list))
+
+    def index(self, paths: List[str]):
+        start_time = perf_counter()
+        discovered = []
+        for path in paths:
+            discovered.extend(self._discover(path))
+
+        logger.info(f"Discovered {len(discovered)} files.", prefix="Discovery")
+
+        self._build_corpus(discovered)
+        self._populate_indices(self.corpus.collect_unprocessed_documents())
+        end_time = perf_counter()
+
+        logger.info(
+            f"{self.corpus.document_count} total files indexed in {end_time - start_time} seconds.",
+            prefix="Index status",
+        )
+
+    def query(self, query: str):
+        start_time = perf_counter()
+        leads = self._trigram_index.query(query)
+        logger.info(
+            f"Narrowed down to {len(leads)} files via trigram search", prefix="Query"
+        )
+        confirmed = []
+        uniques = 0
+        for lead in leads:
+            uid, score = lead
+            lead_path = self.corpus.get_document(uid=uid).key
+            lead_content = ""
+            try:
+                with open(lead_path, "r") as infile:
+                    import mmap
+
+                    m = mmap.mmap(infile.fileno(), 0, prot=mmap.PROT_READ)
+                    lead_content = m.read().decode()
+            except Exception as e:
+                logger.warning(e)
+                logger.warning(f"No content in {lead_path}", prefix="Query")
+
+            results = re.finditer(query, lead_content)
+            hits_in_lead = []
+            for hit in results:
+                start_line, end_line = self._find_line_range(
+                    lead_path, hit.start(), hit.end()
+                )
+                start_offset = self._line_index.query(lead_path)[start_line][0]
+                end_offset = self._line_index.query(lead_path)[end_line][1]
+
+                hits_in_lead.append(
+                    SearchResult(
+                        key=lead_path,
+                        offset_start=start_offset,
+                        offset_end=end_offset,
+                        line_start=start_line,
+                        line_end=end_line,
+                    )
+                )
+
+            if hits_in_lead:
+                confirmed.extend(hits_in_lead)
+                uniques += 1
+        end_time = perf_counter()
+        logger.info(
+            f"{len(confirmed)} hits in {uniques} files ({end_time - start_time} seconds elapsed).",
+            prefix="Query",
+        )
+        return [r.to_dict() for r in confirmed]
+
+    def _discover(self, path_root: str) -> Dict[str, str]:
+        collected = []
+        current = Path(path_root)
+
+        # Avoid any excluded paths
+        if any([current.match(x) for x in self._exclusions]):
+            logger.info(f"{path_root} excluded.", prefix="Discovery")
+            return []
+
+        if current.is_dir():
+            for child_path in current.iterdir():
+                collected.extend(self._discover(str(child_path)))
+
+            return collected
+
+        if current.suffix not in self._file_types:
+            return []
+
+        logger.info(f"Collected {path_root}", prefix="Discovery")
+        return [path_root]
+
+    def _build_corpus(self, discovered: List[str]):
+        total = len(discovered)
+        current = 0
+        for discovered_file in discovered:
+            self.corpus.add_document(key=discovered_file, content="")
+            current += 1
+            logger.info(
+                f"({current}/{total}) Registered {discovered_file} in corpus",
+                prefix="Corpus building",
+            )
+
+    def _populate_indices(self, uids):
+        processes = settings.INDEXING_PROCESSES
+        pool = Pool(processes=processes)
+        chunks = chunkify_content(uids, processes)
+        processed_chunks = pool.map(self._bulk_process, chunks)
+
+        for result in processed_chunks:
+            for uid in result[0]:
+                self._trigram_index.index(
+                    uid.replace("document:", ""), None, None, result[0][uid]
+                )
+            self._line_index._lines.update(result[1])
+
+    # TODO: Tidy up, rethink w.r.t. multiprocessing.
+    def _bulk_process(self, uids: List[str]):
+        trigrams = {}
+        total = len(uids)
+        current = 0
+        for uid in uids:
+            document = self.corpus.get_document(uid=uid)
+            path = document.key
+            try:
+                with open(path, "r") as document_file:
+                    mapped_file = mmap.mmap(
+                        document_file.fileno(), 0, prot=mmap.PROT_READ
+                    )
+                    content = mapped_file.read().decode()
+                    trigrams[uid] = TrigramIndex.trigramize(content)
+                    self._line_index.index(path, content)
+                    current += 1
+                    logger.info(
+                        f"({current}/{total}) Processed {path}", prefix="Indexing"
+                    )
+            except Exception as e:
+                logger.info(e)
+                current += 1
+                logger.warning(
+                    f"({current}/{total}) Could not read {path}, skipping.",
+                    prefix="Indexing",
+                )
+
+        return (trigrams, self._line_index._lines)
+
+    def _find_closest_line(self, path, index):
+        content = self._line_index.query(path)
+
+        for l in content:
+            if content[l][0] <= index <= content[l][1]:
+                return l
+        # TODO: This should not be reachable.
+        return 0
+
+    def _find_line_range(self, key, start, end, padding=5):
+        start_line = self._find_closest_line(key, start)
+        end_line = self._find_closest_line(key, end)
+
+        start_line_range = max(0, start_line - 5)
+        end_line_range = min(len(self._line_index.query(key)) - 1, end_line + 5)
+
+        return (start_line_range, end_line_range)
diff --git a/src/codesearch/line_index.py b/src/codesearch/line_index.py
new file mode 100644
index 0000000..e672d6f
--- /dev/null
+++ b/src/codesearch/line_index.py
@@ -0,0 +1,22 @@
+from .base import IndexBase
+import attr
+
+from .logger import get_logger
+
+logger = get_logger(__name__)
+
+
+@attr.s
+class LineIndex(IndexBase):
+    _lines = attr.ib(default=attr.Factory(dict))
+
+    def index(self, key: str, content: str):
+        self._lines[key] = {}
+        current, count = 0, 0
+        for line in content.split("\n"):
+            self._lines[key][count] = (current, current + len(line))
+            current += len(line)
+            count += 1
+
+    def query(self, key: str):
+        return self._lines[key]
diff --git a/src/codesearch/logger.py b/src/codesearch/logger.py
new file mode 100644
index 0000000..b24d5aa
--- /dev/null
+++ b/src/codesearch/logger.py
@@ -0,0 +1,35 @@
+import logging
+import sys
+import attr
+
+from .colors import highlight
+
+
+@attr.s
+class Logger:
+    logger = attr.ib()
+
+    def info(self, message, prefix=None):
+        prefix_str = ""
+        if prefix:
+            prefix_str = highlight(f"[{prefix}]", "green")
+
+        self.logger.info(f"{prefix_str} {message}")
+
+    def warning(self, message, prefix=None):
+        prefix_str = ""
+        if prefix:
+            prefix_str = highlight(f"[{prefix}]", "yellow")
+
+        self.logger.warning(f"{prefix_str} {message}")
+
+
+def get_logger(name):
+    logger = logging.getLogger(name)
+    logger.setLevel(logging.INFO)
+    handler = logging.StreamHandler(sys.stdout)
+    logger.addHandler(handler)
+
+    logger_obj = Logger(logger=logger)
+
+    return logger_obj
diff --git a/src/codesearch/prefix_tree.py b/src/codesearch/prefix_tree.py
new file mode 100644
index 0000000..3eb2ea0
--- /dev/null
+++ b/src/codesearch/prefix_tree.py
@@ -0,0 +1,66 @@
+import json
+
+import attr
+
+
+@attr.s
+class PrefixTree:
+    root = attr.ib()
+
+    @staticmethod
+    def initialize():
+        root = PrefixTreeNode(value=None)
+        return PrefixTree(root=root)
+
+    def insert(self, value, key, current=None):
+        if current is None:
+            current = self.root
+
+        if not value:
+            current.mappings.append(key)
+            return
+        top = value[0]
+        rest = value[1:]
+
+        next_child = current.children.get(top)
+
+        if next_child:
+            self.insert(rest, key, next_child)
+        else:
+            new_node = PrefixTreeNode(value=top)
+            current.children[top] = new_node
+            self.insert(rest, key, new_node)
+
+    def get(self, value, current=None):
+        if not current:
+            current = self.root
+        if not value:
+            return current.mappings
+
+        top = value[0]
+        rest = value[1:]
+
+        next_child = current.children.get(top)
+
+        if next_child:
+            return self.get(rest, next_child)
+
+    def to_dict(self):
+        return self.root.to_dict()
+
+    def to_json(self):
+        return json.dumps(self.to_dict())
+
+
+@attr.s
+class PrefixTreeNode:
+    value = attr.ib()
+    mappings = attr.ib(default=attr.Factory(list))
+    children = attr.ib(default=attr.Factory(dict))
+
+    def to_dict(self):
+        return {
+            "value": self.value,
+            "mappings": self.mappings,
+            "children": [child.to_dict() for child in self.children.values()],
+        }
diff --git a/src/codesearch/process_utils.py b/src/codesearch/process_utils.py
new file mode 100644
index 0000000..e5d6613
--- /dev/null
+++ b/src/codesearch/process_utils.py
@@ -0,0 +1,15 @@
+def chunkify_content(content, chunk_count, chunk_size=None):
+    if chunk_size is None:
+        chunk_size = int(len(content) / chunk_count)
+    chunks = []
+    last_boundary = 0
+
+    for i in range(chunk_count):
+        if i == chunk_count - 1:
+            chunks.append(content[last_boundary:])
+        else:
+            chunks.append(content[last_boundary : last_boundary + chunk_size])
+
+        last_boundary += chunk_size
+
+    return chunks
diff --git a/src/codesearch/server.py b/src/codesearch/server.py
new file mode 100644
index 0000000..bebf1b7
--- /dev/null
+++ b/src/codesearch/server.py
@@ -0,0 +1,81 @@
+import json
+import socket
+import pyinotify
+import attr
+from codesearch.watcher import WatchHandler
+from codesearch.indexer import Indexer
+from codesearch.constants import QUERY_STRING_LENGTH
+from pathlib import Path
+from codesearch.settings import settings
+
+from codesearch.logger import get_logger
+
+logger = get_logger(__name__)
+
+
+@attr.s
+class Server:
+    indexer = attr.ib()
+    watched = attr.ib()
+    _notifier = attr.ib(default=None)
+    _socket = attr.ib(default=None)
+
+    def _handle_socket(self, *, socket):
+        socket.bind((settings.SOCKET_HOST, settings.SOCKET_PORT))
+        socket.listen()
+
+        logger.info(
+            f"Listening on {settings.SOCKET_HOST}:{settings.SOCKET_PORT}",
+            prefix="Server",
+        )
+
+        while True:
+            conn, _ = socket.accept()
+            query_string = conn.recv(QUERY_STRING_LENGTH).decode()
+            logger.info(f"Query string: {query_string}", prefix="Query")
+            if query_string:
+                try:
+                    query_results = self.indexer.query(query_string)
+                    response = json.dumps(query_results).encode()
+                    response_length = str(len(response.decode()))
+                    with open(Path(settings.BUFFER_PATH).expanduser(), "wb") as outfile:
+                        outfile.write(response)
+                    conn.sendall(response_length.encode())
+                except KeyboardInterrupt:
+                    raise e
+                except Exception as e:
+                    logger.warning(e)
+                    pass
+
+    def _start_socket(self):
+        try:
+            with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as socket_obj:
+                self._socket = socket_obj
+                self._handle_socket(socket=socket_obj)
+        except Exception as e:
+            logger.warning(e)
+            raise e
+
+    def _start_watch(self):
+        watch_manager = pyinotify.WatchManager()
+
+        for path in self.watched:
+            logger.info(f"Watching {path}", prefix="Server")
+            watch_manager.add_watch(path, pyinotify.ALL_EVENTS, rec=True)
+
+        event_handler = WatchHandler(indexer=self.indexer)
+        notifier = pyinotify.ThreadedNotifier(watch_manager, event_handler)
+        notifier.start()
+        self._notifier = notifier
+
+    def run(self):
+        collected = {}
+
+        self.indexer.index(self.watched)
+
+        try:
+            self._start_watch()
+            self._start_socket()
+        except:
+            self._socket.close()
+            self._notifier.stop()
diff --git a/src/codesearch/settings.py b/src/codesearch/settings.py
new file mode 100644
index 0000000..ea5692a
--- /dev/null
+++ b/src/codesearch/settings.py
@@ -0,0 +1,45 @@
+import json
+
+from pathlib import Path
+import attr
+
+from .constants import SETTINGS_KEYS
+
+SETTINGS_PATH = "~/.codesearchrc"
+
+default_settings = {
+    "SOCKET_HOST": "127.0.0.1",
+    "SOCKET_PORT": 65126,
+    "EXCLUDES": [],
+    "FILE_TYPES": [],
+    "SIGNIFICANCE_THRESHOLD": 0,
+    "WATCHED": [],
+    "INDEXING_PROCESSES": 4,
+    "BUFFER_PATH": "~/.codesearchbuffer",
+}
+
+
+@attr.s
+class Settings:
+    settings = attr.ib(default=attr.Factory(dict))
+
+    def from_file(self, path: str):
+        settings_path = Path(SETTINGS_PATH).expanduser()
+
+        if not settings_path.exists():
+            self.settings = default_settings
+            return
+
+        with open(path, "r") as settings_file:
+            self.settings = json.load(settings_file)
+
+    def __getattr__(self, key):
+        if key not in SETTINGS_KEYS:
+            raise KeyError(f"{key} not a valid settings property")
+
+        return self.settings[key]
+
+
+settings = Settings()
+
+settings.from_file(Path(SETTINGS_PATH).expanduser())
diff --git a/src/codesearch/test_indexer.py b/src/codesearch/test_indexer.py
new file mode 100644
index 0000000..e0d3960
--- /dev/null
+++ b/src/codesearch/test_indexer.py
@@ -0,0 +1,49 @@
+import pytest
+
+from .indexer import Indexer
+
+
+@pytest.fixture()
+def indexer():
+    return Indexer()
+
+
+def test_indexer_builds_trigram_set_for_given_document(indexer):
+    mock_document = "now that's a doc"
+    mock_path = "/home/documents/cool_doc"
+
+    indexer.index(path=mock_path, content=mock_document)
+
+    expected_trigrams = [
+        "now",
+        "ow ",
+        "w t",
+        " th",
+        "tha",
+        "hat",
+        "at'",
+        "t's",
+        "'s ",
+        "s a",
+        " a ",
+        "a d",
+        " do",
+        "doc",
+    ]
+
+    assert indexer.trigrams == {mock_path: set(expected_trigrams)}
+
+
+def test_indexer_preserves_previous_trigram_sets_on_index(indexer):
+    mock_document_1 = "wow"
+    mock_document_2 = "woa"
+    mock_path_1 = "/home"
+    mock_path_2 = "/somewhere_else"
+
+    indexer.index(path=mock_path_1, content=mock_document_1)
+
+    assert indexer.trigrams == {mock_path_1: set(["wow"])}
+
+    indexer.index(path=mock_path_2, content=mock_document_2)
+
+    assert indexer.trigrams == {mock_path_1: set(["wow"]), mock_path_2: set(["woa"])}
diff --git a/src/codesearch/test_prefix_tree.py b/src/codesearch/test_prefix_tree.py
new file mode 100644
index 0000000..2d52ac6
--- /dev/null
+++ b/src/codesearch/test_prefix_tree.py
@@ -0,0 +1,67 @@
+import pytest
+
+from .prefix_tree import PrefixTree
+
+
+@pytest.fixture
+def prefix_tree():
+    return PrefixTree.initialize()
+
+
+def test_base_tree_has_a_root_node(prefix_tree, snapshot):
+    assert prefix_tree.to_dict() == snapshot
+
+
+def test_insert_single_string(prefix_tree, snapshot):
+    mock_value = "abc"
+    mock_key = "key_1"
+    prefix_tree.insert(value=mock_value, key=mock_key)
+    assert prefix_tree.to_dict() == snapshot
+    assert prefix_tree.get(value=mock_value) == [mock_key]
+
+
+def test_insert_single_character_(prefix_tree, snapshot):
+    mock_value = "a"
+    mock_key = "key_1"
+    prefix_tree.insert(value=mock_value, key=mock_key)
+    assert prefix_tree.to_dict() == snapshot
+    assert prefix_tree.get(value=mock_value) == [mock_key]
+
+
+def test_insert_overlapping_strings(prefix_tree, snapshot):
+    mock_value_1 = "abcd"
+    mock_key_1 = "key_1"
+    mock_value_2 = "abce"
+    mock_key_2 = "key_2"
+    prefix_tree.insert(value=mock_value_1, key=mock_key_1)
+    prefix_tree.insert(value=mock_value_2, key=mock_key_2)
+    assert prefix_tree.to_dict() == snapshot
+    assert prefix_tree.get(value=mock_value_1) == [mock_key_1]
+    assert prefix_tree.get(value=mock_value_2) == [mock_key_2]
+
+
+def test_insert_multiple_keys_same_string(prefix_tree, snapshot):
+    mock_value = "abcd"
+    mock_key_1 = "key_1"
+    mock_key_2 = "key_2"
+    prefix_tree.insert(value=mock_value, key=mock_key_1)
+    prefix_tree.insert(value=mock_value, key=mock_key_2)
+    assert prefix_tree.to_dict() == snapshot
+    assert prefix_tree.get(value=mock_value) == [mock_key_1, mock_key_2]
+
+
+def test_insert_strings_subsets_of_each_other(prefix_tree, snapshot):
+    mock_value_1 = "abcd"
+    mock_key_1 = "key_1"
+    mock_value_2 = "abc"
+    mock_key_2 = "key_2"
+    prefix_tree.insert(value=mock_value_1, key=mock_key_1)
+    prefix_tree.insert(value=mock_value_2, key=mock_key_2)
+    assert prefix_tree.to_dict() == snapshot
+    assert prefix_tree.get(value=mock_value_1) == [mock_key_1]
+    assert prefix_tree.get(value=mock_value_2) == [mock_key_2]
+
+
+def test_serializes_to_json(prefix_tree, snapshot):
+    prefix_tree.insert(value="abcd", key="key_1")
+    assert prefix_tree.to_json() == snapshot
diff --git a/src/codesearch/trigram_index.py b/src/codesearch/trigram_index.py
new file mode 100644
index 0000000..a147544
--- /dev/null
+++ b/src/codesearch/trigram_index.py
@@ -0,0 +1,47 @@
+from typing import List, Optional
+
+import attr
+from .settings import settings
+from .base import IndexBase
+from .prefix_tree import PrefixTree
+
+
+@attr.s
+class TrigramIndex(IndexBase):
+    _threshold = attr.ib(default=settings.SIGNIFICANCE_THRESHOLD)
+    _tree = attr.ib(attr.Factory(PrefixTree.initialize))
+
+    def index(self, uid, key: str, content: str, trigrams):
+        if content:
+            trigrams = TrigramIndex.trigramize(content)
+
+        for trigram in trigrams:
+            self._tree.insert(trigram, uid)
+
+    def query(self, query: str, haystack: Optional[List[str]] = None) -> List[str]:
+        query_trigrams = TrigramIndex.trigramize(query)
+        results = {}
+
+        for trigram in query_trigrams:
+            result_set = self._tree.get(trigram)
+            if result_set:
+                results[trigram] = result_set
+
+        matches = {}
+
+        for result in results:
+            for doc in results[result]:
+                matches[doc] = matches.get(doc, 0) + 1
+
+        significant_results = []
+        for uid, occurrences in matches.items():
+            score = occurrences / len(query_trigrams)
+            if score >= self._threshold:
+                significant_results.append((f"document:{uid}", score))
+
+        significant_results.sort(reverse=True, key=lambda x: x[0])
+        return significant_results
+
+    @staticmethod
+    def trigramize(content: str) -> List[str]:
+        return {content[pos : pos + 3].lower() for pos in range(len(content) - 2)}
diff --git a/src/codesearch/watcher.py b/src/codesearch/watcher.py
new file mode 100644
index 0000000..92c25db
--- /dev/null
+++ b/src/codesearch/watcher.py
@@ -0,0 +1,14 @@
+import pyinotify
+import attr
+
+from .logger import get_logger
+
+logger = get_logger(__name__)
+
+
+@attr.s
+class WatchHandler(pyinotify.ProcessEvent):
+    indexer = attr.ib()
+
+    def process_IN_MODIFY(self, event):
+        self.indexer.index([event.pathname])
diff --git a/src/setup.py b/src/setup.py
new file mode 100644
index 0000000..6d1bf11
--- /dev/null
+++ b/src/setup.py
@@ -0,0 +1,9 @@
+from setuptools import setup
+
+setup(
+    name="codesearch",
+    version="0.1",
+    packages=["codesearch"],
+    install_requires=["pyinotify", "attr"],
+    entry_points={"console_scripts": ["codesearch=codesearch.cli:main"]},
+)
diff --git a/tasks.py b/tasks.py
new file mode 100644
index 0000000..00ca26b
--- /dev/null
+++ b/tasks.py
@@ -0,0 +1,6 @@
+from invoke import task
+
+
+@task
+def lint(ctx):
+    ctx.run("black *.py src")